DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. CHINA 202110117226.X filed on January 28, 2021.
Claim Objections
Claims 1 and 11 are objected to because of the following informalities: ‘… wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm,’ where ‘bases on’ should read ‘based on’. Appropriate correction is required.
Response to Amendment
The Amendment filed July 7, 2026, has been entered. Claims 1- 20 remain pending in the application.
Response to Arguments:
In reference to Applicant’s arguments:
-Claim rejections under 35 U.S.C. 101.
Examiner’s response:
Regarding applicant’s arguments of 35 U.S.C. 101 rejections on remarks page 8 paragraphs 2-3, stating that “Application respectfully submits that the amended claims do not recite the "performing operations research and optimization calculation…to obtain a calculation result" step that is alleged by the Office Action as a mathematical concept.” The examiner submits that the argument is fully considered, however, while amended claims 1 and 11 did not recite the element “performing operations research and optimization calculation…to obtain a calculation result”, the dependent claims such as claim 5 and 15 still recite “…by using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result”, which is considered a mathematical relationship, mathematical formula or equation, or mathematical calculation. For details, see the specification in paragraphs [0110-0113] state “the goodness of the calculation result indicates a degree of the calculation result of the operations research and optimization algorithm closing to the optimal solution. For example, a difference η may reflect the goodness of the calculation result” and that the calculation result is obtained by a formula:
PNG
media_image1.png
80
242
media_image1.png
Greyscale
In the foregoing formula, x represents a calculation result obtained by performing operations research and optimization calculation on the data of the current application scenario, and s represents a relaxation solution obtained by performing operations research and optimization calculation on the data of the current application scenario.” Here, an optimized solution or optimized calculation result is expressed with an equation, see MPEP 2106.04(a)(2), subsection I . Further, there are some limitations from claim 3 such as "analyzing the feature of the data" and "determining that the data of the current application scenario is abnormal data", as well as some limitations of claim 4 such as "analyzing the result of the current application scenario" or "determining that the data of the current application scenario is abnormal data" still recite mental processes since a person can perform these mentally. Since claims 5-7 depend on claim 3, these dependent claims are also rejected under 35 U.S.C. 101. Further, claim 10 recites "and determining the operations research and optimization algorithm based on the task type," which is a mental process. Additionally, from page 8, paragraphs 2-3 of the applicant’s remarks, examiner was unable to find the improvement that connect the invention to a practical application since this was not stated in the remarks. Therefore, the rejections under 35 U.S.C. 101 are held. See 35 U.S.C. 101 rejections for more details.
In reference to Applicant’s arguments:
-Claim rejections under 35 U.S.C. 103.
Examiner’s response:
Regarding applicant’s arguments on page 9 stating “Applicant respectfully traverses these rejections for the comments set forth below. Fu has not been shown to teach or suggest that "the hyperparameter inference model is obtained through dynamic training based on [both] training data obtained in a historical application scenario and training data obtained in the current application scenario”, and from applicant’s remarks from pages 9-10 of “Applicant respectfully submits that Fu merely refers to inputting the data obtained from the past 14 days into the model to obtain the predicted energy consumption value for the 15th day. In contrast to the claims, Fu has not been shown to teach or suggest that the model is obtained through dynamic training based on the data obtained from the past 14 days and the predicted energy consumption value for the 15th day. Therefore, Applicant respectfully submits that Fu has not been shown to teach or suggest that "the hyperparameter inference model is obtained through dynamic training based on [both] training data obtained in a historical application scenario and training data obtained in the current application scenario,” and the examiner submits the applicant's arguments with respect to the rejections of claims 1 and 11, and their respective dependent claims under 35 U.S.C. 103 in the previous office action have been fully considered, but are moot because the arguments do not apply to the combination of references used in the current rejection and because the new ground of rejection does not rely on the Gao reference (Gao, X. et al., in US PG Pub. No. US20200410299A1, published on December 31, 2020) applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Regarding the applicant’s remarks on page 11, paragraph 2, stating “Fu has not been shown to teach or suggest "sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface,” particularly where “the result of the current application scenario is obtained bases on [(1)] the data of the current application scenario, [(2)] the hyperparameter, and [(3)] the operations research and optimization algorithm" as recited in amended claim 1”, the examiner submits the applicant's arguments with respect to the limitation that the reference Fu did not teach the elements mentioned on page 11, paragraph 2, however, the reference Achin teaches these elements, see Achin in paragraph [0235] describe “Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return.” Here, ACHIN shows that for each data file, the user can upload this directly to an API or user interface. The user also receives calculated predictions or results from the API interface.
Further, see ACHIN describe in [0293] that “(6) External system 660. As with any other Internet application, the use of APIs may allow external systems to integrate with the predictive modeling system 100 at any layer of the architecture 600. For example, a business dashboard application can access graphical visualization and modeling results through the interface services layer 620...” Here, ACHIN shows the results from a live business application (i.e. current application scenario) are sent through the API, and from [0235] the results from the API can be sent to a user.
Further, see ACHIN in [0161] mention “Returning to FIG. 3, at step 340 of method 300, a result of executing the selected modeling procedure according to the resource allocation schedule may be received. These results may include one or more predictive models generated by the executed modeling procedure. In some embodiments, the execution of the modeling procedure may include fitting the prediction model to one or more datasets associated with the prediction problem, so that the prediction model received in step 340 may include the prediction problem Fitted to the dataset associated with. Fitting the predictive model to the data set of the predictive problem includes adjusting one or more hyperparameters of a predictive modeling procedure that generates the predictive model, adjusting one or more parameters of the generated predictive model.” Here, ACHIN shows the information that the current application scenario of ‘fitting the prediction model to one or more datasets associated with the prediction problem’, is based on adjusting one or more hyperparameters (i.e. hyperparameter) of a predictive modeling procedure (i.e. operations research and optimization algorithm). See section 35 U.S.C. 103 rejections for details.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefore, subject to the conditions and requirements of this title.
Claims 3-7, 10, 13-17, and 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea (mental process or math concept) without significantly more.
Claim 3:
Regarding claim 3, in step 1 of the 101-analysis set forth in MPEP 2106, the claim recites
“The method according to claim 1, wherein the method further comprises: analyzing the feature of the data; and determining that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data, a user weight preference parameter in the data, or a problem structure parameter of the data” , and a method is one of the four statutory categories of invention.
In step 2A prong 1 of the 101-analysis set forth in the MPEP 2106, the examiner has determined that the following limitations recite a process that, under the broadest reasonable interpretation, covers a mental process but for recitation of generic computer components:
The method according to claim 1, wherein the method further comprises: analyzing the feature of the data; (this is considered a mental process, since a person can mentally evaluate and analyze a data feature, see MPEP 2106.04(a)(2)(III)),
and determining that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data, a user weight preference parameter in the data, or a problem structure parameter of the data, (this is considered a mental process, since a person can mentally evaluate and determine if data has abnormal data, see MPEP 2106.04(a)(2)(III)),
If claim limitations, under their broadest reasonable interpretation, cover performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 1 do not integrate this judicial exception into a practical application:
An operations research and optimization method, comprising: obtaining, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
obtaining, by the computing device, a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
and sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea.
In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, which is not indicative of significantly more, and additional elements iii, iv, and vi recite mere data gathering, are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)),
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim 4:
Regarding claim 4, it is dependent upon claim 1, and thereby incorporates the limitations of, and corresponding analysis applied to claim 1. Further, claim 4 recites the following abstract ideas:
The method according to claim 1, wherein the method further comprises: analyzing the result of the current application scenario; (this recites a mental process, since a person can mentally evaluate and analyze a result, see MPEP 2106.04(a)(2)(III)),
and when the result of the current application scenario does not meet a preset condition, determining that the data of the current application scenario is abnormal data, (this is considered a mental process, since a person can mentally evaluate and determine if a result does not meet a preset condition, this means the data has abnormal data, see MPEP 2106.04(a)(2)(III)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 1 do not integrate this judicial exception into a practical application:
An operations research and optimization method, comprising: obtaining, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
obtaining, by the computing device, a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
and sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea.
In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, which is not indicative of significantly more, and additional elements iii, iv, and vi recite mere data gathering, are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)),
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim 5:
Regarding claim 5, it is dependent upon claim 3, and thereby incorporates the limitations of, and corresponding analysis applied to claim 3. Further, claim 5 recites the following abstract idea:
The method according to claim 3, wherein the method further comprises: optimizing the hyperparameter of the operations research and optimization algorithm by using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result, (This is considered a mathematical relationship, mathematical formula or equation, or mathematical calculation, see the specification in paragraphs [0110-0113] state “the goodness of the calculation result indicates a degree of the calculation result of the operations research and optimization algorithm closing to the optimal solution. For example, a difference η may reflect the goodness of the calculation result” and that the calculation result is obtained by a formula:
PNG
media_image1.png
80
242
media_image1.png
Greyscale
In the foregoing formula, x represents a calculation result obtained by performing operations research and optimization calculation on the data of the current application scenario, and s represents a relaxation solution obtained by performing operations research and optimization calculation on the data of the current application scenario.” Here, an optimized solution or optimized calculation result is expressed with an equation, see MPEP 2106.04(a)(2), subsection I),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mathematical concept but for the recitation of generic computer components, then it falls within the mathematical concept grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 6:
Regarding claim 6, it is dependent upon claim 5, and thereby incorporates the limitations of, and corresponding analysis applied to claim 5. Further, claim 6 recites the following additional elements:
The method according to claim 5, wherein the method further comprises: recording the abnormal data and the optimized calculation result corresponding to the abnormal data into a training data set, (In step 2A, prong 2, this recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g),). In step 2B, this insignificant extra-solution activity is well understood routine and conventional activity which includes receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) – see MPEP 2106.05(d) (II)(i),
…used to train the hyperparameter inference model, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 7:
Regarding claim 7, it is dependent upon claim 6, and thereby incorporates the limitations of, and corresponding analysis applied to claim 6. Further, claim 7 recites the following abstract idea:
The method according to claim 6, wherein the method further comprises: determining that the hyperparameter inference model is to be updated; (this is considered a mental process, since a person can mentally evaluate and determine if the hyperparameter inference model is to be updated, see MPEP 2106.04(a)(2)(III)),
Further, claim 7 recites the following additional element:
and training the hyperparameter inference models, based on training data in the training data set, to obtain an updated hyperparameter inference model, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea.
Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 10:
Regarding claim 10, it is dependent upon claim 1, and thereby incorporates the limitations of, and corresponding analysis applied to claim 1. Further, claim 10 recites the following abstract idea:
and determining the operations research and optimization algorithm based on the task type, (this is considered a mental process, since a person can mentally evaluate and determine an algorithm based on a task type, see MPEP 2106.04(a)(2)(III)),
Further, claim 10 recites the following additional element:
The method according to claim 1, wherein the method further comprises: obtaining an operations research and optimization task type configured by the user; (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea.
In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 1 do not integrate this judicial exception into a practical application:
An operations research and optimization method, comprising: obtaining, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
obtaining, by the computing device, a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
and sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea.
In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, which is not indicative of significantly more, and additional elements iii, iv, and vi recite mere data gathering, are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)), Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 13:
Regarding claim 13, in step 1 of the 101-analysis set forth in MPEP 2106, the claim recites
“The computing device according to claim 11, wherein the programming instructions are for execution by the at least one processor to: analyze the feature of the data; and determine that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data, a user weight preference parameter in the data, or a problem structure parameter of the data” , and a device or system is one of the four statutory categories of invention.
In step 2A prong 1 of the 101-analysis set forth in the MPEP 2106, the examiner has determined that the following limitations recite a process that, under the broadest reasonable interpretation, covers a mental process but for recitation of generic computer components:
The computing device according to claim 11, wherein the programming instructions are for execution by the at least one processor to: analyze the feature of the data; (this is considered a mental process, since a person can mentally evaluate and analyze a data feature, see MPEP 2106.04(a)(2)(III)),
and determine that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data, a user weight preference parameter in the data, or a problem structure parameter of the data, (this is considered a mental process, since a person can mentally evaluate and determine if data has abnormal data, see MPEP 2106.04(a)(2)(III)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea.
If claim limitations, under their broadest reasonable interpretation, cover performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 11 do not integrate this judicial exception into a practical application:
A computing device, comprising at least one processor and one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor… (In step 2A, prong 2, this is considered a generic computer component being used as a tool. – see MPEP 2106.05(f)),
to: obtain, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
obtain a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
and send a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea. In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, and additional element iii recites a generic computer component being used as a tool, which are not indicative of significantly more. The additional elements iv, vi, and vii recite mere data gathering, and are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)),
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim 14:
Regarding claim 14, it is dependent upon claim 11, and thereby incorporates the limitations of, and corresponding analysis applied to claim 11. Further, claim 14 recites the following abstract ideas:
The computing device according to claim 11, wherein the programming instructions are for execution by the at least one processor to: analyze the result of the current application scenario; (this is considered a mental process, since a person can mentally evaluate and analyze a calculation result, see MPEP 2106.04(a)(2)(III)),
and when the result of the current application scenario does not meet a preset condition, determine that the data of the current application scenario is abnormal data, (this is considered a mental process, since a person can mentally evaluate and determine if a calculation result does not meet a preset condition, this means the data has abnormal data, see MPEP 2106.04(a)(2)(III)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 11 do not integrate this judicial exception into a practical application:
A computing device, comprising at least one processor and one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor… (In step 2A, prong 2, this is considered a generic computer component being used as a tool. – see MPEP 2106.05(f)),
to: obtain, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
obtain a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
and send a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea.
In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, and additional element iii recites a generic computer component being used as a tool, which are not indicative of significantly more. The additional elements iv, vi, and vii recite mere data gathering, and are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)),
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim 15:
Regarding claim 15, it is dependent upon claim 13, and thereby incorporates the limitations of, and corresponding analysis applied to claim 13. Further, claim 15 recites the following abstract idea:
The computing device according to claim 13, wherein the programming instructions are for execution by the at least one processor to: optimize the hyperparameter of the operations research and optimization algorithm by using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result, (This is considered a mathematical relationship, mathematical formula or equation, or mathematical calculation, see the specification in paragraphs [0110-0113] state “the goodness of the calculation result indicates a degree of the calculation result of the operations research and optimization algorithm closing to the optimal solution. For example, a difference η may reflect the goodness of the calculation result” and that the calculation result is obtained by a formula:
PNG
media_image1.png
80
242
media_image1.png
Greyscale
In the foregoing formula, x represents a calculation result obtained by performing operations research and optimization calculation on the data of the current application scenario, and s represents a relaxation solution obtained by performing operations research and optimization calculation on the data of the current application scenario.” Here, an optimized solution or optimized calculation result is expressed with an equation, see MPEP 2106.04(a)(2), subsection I),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mathematical concept but for the recitation of generic computer components, then it falls within the mathematical concept grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 16:
Regarding claim 16, it is dependent upon claim 15, and thereby incorporates the limitations of, and corresponding analysis applied to claim 15. Further, claim 16 recites the following additional elements:
The computing device according to claim 15, wherein the programming instructions are for execution by the at least one processor to: record the abnormal data and the optimized calculation result corresponding to the abnormal data into a training data set, (In step 2A, prong 2, this recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g),). In step 2B, this insignificant extra-solution activity is well understood routine and conventional activity which includes receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) – see MPEP 2106.05(d) (II)(i),
…used to train the hyperparameter inference model, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 17:
Regarding claim 17, it is dependent upon claim 16, and thereby incorporates the limitations of, and corresponding analysis applied to claim 16. Further, claim 17 recites the following abstract idea:
The computing device according to claim 16, wherein the programming instructions are for execution by the at least one processor to: determine that the hyperparameter inference model is to be updated; (this is considered a mental process, since a person can mentally evaluate and determine if the hyperparameter inference model is to be updated, see MPEP 2106.04(a)(2)(III)),
Further, claim 17 recites the following additional element:
and train the hyperparameter inference model, based on training data in the training data set, to obtain an updated hyperparameter inference model, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Claim 20:
Regarding claim 20, it is dependent upon claim 11, and thereby incorporates the limitations of, and corresponding analysis applied to claim 11. Further, claim 20 recites the following abstract idea:
and determine the operations research and optimization algorithm based on the task type, (this is considered a mental process, since a person can mentally evaluate and determine an algorithm based on a task type, see MPEP 2106.04(a)(2)(III)),
Further, claim 20 recites the following additional element:
The computing device according to claim 11, wherein the programming instructions are for execution by the at least one processor to: obtain an operations research and optimization task type configured by the user; (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)), (In step 2B, this is also considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it falls within the mental process grouping of abstract ideas. Accordingly, the claim “recites” an abstract idea. In step 2A prong 2 of the 101-analysis set forth in MPEP 2106, the examiner has determined that the following additional elements from independent claim 11 do not integrate this judicial exception into a practical application:
A computing device, comprising at least one processor and one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor… (In step 2A, prong 2, this is considered a generic computer component being used as a tool. – see MPEP 2106.05(f)),
to: obtain, by a computing device, data of a current application scenario and a feature of the data, (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface, (In step 2A, prong 2, this is considered mere instructions to apply an exception using generic computer – see MPEP 2106.05(f)),
obtain a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models, wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; (In step 2A, prong 2, obtaining recites mere data gathering, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
and send a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm, (In step 2A, prong 2, sending recites mere data transmitting, which is considered insignificant extra-solution activity – see MPEP 2106.05(g)),
Since the claim as a whole, looking at the additional elements individually and in combination, does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea. In step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, additional element v recites mere instructions to apply the judicial exception using generic computer components, and additional element iii recites a generic computer component being used as a tool, which are not indicative of significantly more. The additional elements iv, vi, and vii recite mere data gathering, and are considered insignificant extra-solution activities. In step 2B, these insignificant extra-solution activities are well understood routine and conventional activities which include receiving or transmitting data over a network from court case Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016), – see MPEP 2106.05(d) (II)(i)),
Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 2, 3, 8, 9, 11, 12, 13, 18, and 19 are rejected under 35 U.S.C. 103 as being unpatentable by Fu, Q. et al., provided in the January 6, 2025 IDS, (Pub. No. CN 111178626A), published May 19, 2020, (hereafter, FU) in view of Achin, J. et al., (Pub. No. JP2019537125A), published on December 19, 2019, (hereafter, ACHIN), further in view of Gao, X. et al., in US PG Pub. No. US20200410299A1, published on December 31, 2020, (hereafter, Gao).
Claim 1:
Regarding claim 1, FU teaches “An operations research and optimization method, comprising: obtaining, by a computing device, data of a current application scenario and a feature of the data …”
See FU in [0013] describe “The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms.” Here, FU discloses obtaining energy consumption data (i.e. obtain data of a current application scenario) and extract high-level features (i.e. and a feature of the data).
Further, see FU in [0103] describe “A building energy consumption monitoring and prediction system includes: a real-time energy consumption data acquisition module, an environmental data acquisition module, a data communication module, an information processing module, a server, a display, a human-computer interaction module, and an information storage module;” Here, FU mentions a monitoring and prediction system (i.e. computing device) that acquires data and related information.
obtaining, by the computing device, a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference models…
See FU describe in [0137] "that after obtaining a new set of hyperparameters, the values of the hyperparameters in the GAN are updated, and the next GAN training is started. This process is repeated n times, for example, 10 times, until the hyperparameters no longer change. At this point, the optimal GAN prediction model can be obtained."
Further, FU describes in [0139] "The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, and finally a prediction model is obtained to predict building energy consumption." This shows that FU discloses obtaining a new set of hyperparameters in [0137] that uses reinforcement learning algorithm to optimize the hyperparameters for the general adversarial network model (i.e. hyperparameter inference model), and in [0139] all the features are then used as input for model training, then the hyperparameters are also optimized for the same model (i.e. obtain a hyperparameter of an operations research and optimization algorithm based on feature of data).
Further, see FU in [0107] mention “the server stores at least one executable instruction that enables the server to perform operations corresponding to the building energy consumption prediction method”, where FU shows this performs operations that relates to building a prediction method for energy consumption (i.e. part of an operations research and optimization algorithm). In the specification paragraph [0003], which states that an operations research and optimization algorithm have “methods to realize effective management, correct decision-making and modern management. An operations research and optimization algorithm is widely used in daily life and production practice... the operations research and optimization algorithm is indispensable and irreplaceable in many industries and fields such as social networking, entertainment, education, transportation, security, industry, logistics, and e-commerce.” This states that an operations research and optimization algorithm is involved in applications to various fields and production such as the example FU notes in building a method for predicting energy consumption. Further, see FU in [0103] regarding the computing device similar to above.
However, FU did not teach “…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface;”
or “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario; or “and sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm.
In an analogous art, ACHIN teaches “…wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface;
See ACHIN in [0235] describe “Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return.” Here, ACHIN shows that for each data file, the user can upload this directly to an API or user interface.
Further, ACHIN teaches “and sending, by the computing device, a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm”
See ACHIN in [0235] describe “Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return.” Here, ACHIN shows that for each data file, the user can upload this directly to an API or user interface. The user also receives calculated predictions or results from the API interface.
Further, see ACHIN describe in [0293] that “(6) External system 660. As with any other Internet application, the use of APIs may allow external systems to integrate with the predictive modeling system 100 at any layer of the architecture 600. For example, a business dashboard application can access graphical visualization and modeling results through the interface services layer 620. An external data warehouse or even a live business application can provide the modeled dataset to the analytics service layer 630 through a data integration platform. The reporting application can access all modeling results from a particular time period through the data services layer 650.” Here, ACHIN shows the results from a live business application (i.e. current application scenario) are sent through the API, and from [0235] the results from the API can be sent to a user.
Further, see ACHIN in [0161] mention “Returning to FIG. 3, at step 340 of method 300, a result of executing the selected modeling procedure according to the resource allocation schedule may be received. These results may include one or more predictive models generated by the executed modeling procedure. In some embodiments, the execution of the modeling procedure may include fitting the prediction model to one or more datasets associated with the prediction problem, so that the prediction model received in step 340 may include the prediction problem Fitted to the dataset associated with. Fitting the predictive model to the data set of the predictive problem includes adjusting one or more hyperparameters of a predictive modeling procedure that generates the predictive model, adjusting one or more parameters of the generated predictive model.” Here, ACHIN shows the information that the current application scenario of ‘fitting the prediction model to one or more datasets associated with the prediction problem’, is based on adjusting one or more hyperparameters (i.e. hyperparameter) of a predictive modeling procedure (i.e. operations research and optimization algorithm).
Later, see ACHIN mention in [0223] “To support the breadth of user requirements, some embodiments of the user interface 120 provide a set of interface tools that reflect the model building process. Each tool may also provide a series of features from basic to advanced. The first step in the model building process may involve loading and preparing a dataset. As discussed previously, a user may upload files or define how to access data from online systems. In the context of a modeling project group or hierarchy, the user may also specify which part of the parent dataset is to be used for the current project and which parts will be added." Note, examiner construes current application scenario to mean a specific context or situation within which a machine learning model or system is applied to solve a particular problem or make predictions. Here, ACHIN mentions the user may determine the dataset to use for a current project, which in this case is a current application scenario for which a user uploads to run models to solve problems or make predictions.
See ACHIN in [0242] describe for more details. Here, ACHIN shows that the results of a prediction modeling system 100 are stored until needed.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research using an application programming interface.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
However, FU in view of ACHIN did not teach “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario;”
In an analogous art, Gao teaches “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario;”
See Gao in [0070] describe "All these experiments show that the proposed novel class incremental learning method, SupportNet, solves the catastrophic forgetting problem by combining the strength of deep learning and SVM. SupportNet can efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. With the help of two powerful consolidation regularizers, the support data can effectively help the deep learning model prevent the catastrophic forgetting issue, eliminate the necessity of retraining the model from scratch, and maintain a stable learned representation that corresponds to the old and the new data." Here, Gao shows that the model called SupportNet can ‘efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information’ to indicate training uses both old data (i.e. data from historical application scenario) and new data (i.e. data obtained in the current application scenario).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the references of FU and ACHIN, with the teachings of Gao by using the teachings of FU and ACHIN of optimizing a model with feature analysis, with Gao’s teaching of dynamic training using historical data and current application data.
One of ordinary skill in the art would be motivated to do so because by integrating Gao’s framework into the methods of FU and ACHIN, one with ordinary skill in the art would achieve a method like “SupportNet can efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. With the help of two powerful consolidation regularizers, the support data can effectively help the deep learning model prevent the catastrophic forgetting issue, eliminate the necessity of retraining the model from scratch, and maintain a stable learned representation that corresponds to the old and the new data,” (see Gao in [0070]), and “the present method is a highly effective framework (called SupportNet in the following), which can perform class incremental learning without catastrophic forgetting,” (see Gao in [0055]).
Claim 2:
Regarding claim 2, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 1.
Further, FU teaches “The method according to claim 1, wherein the obtaining a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference model comprises: inputting the feature of the data to the hyperparameter inference model;”
See FU describe in [0139] "The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, and finally a prediction model is obtained to predict building energy consumption." Here, FU mentions from [0139] that after extracting the features, the features are then used as input for training the model. Then, the hyperparameters are optimized using the same model.
Further, see FU in [0189] describe " the generator employs a Long Short-Term Memory (LSTM) neural network initialized with Xavier. The LSTM input consists of collected and generated energy consumption features, which are then processed through 500 hidden units and converted into a single output as the predicted energy consumption value. The sequence step size parameter is set to 14." Here, FU further specifies that the LSTM input consists of energy consumption features (i.e. inputting the feature of data) to the generative adversarial network (GAN) model (i.e. hyperparameter inference model), See FU in [0170-0171] for more information.
Further, FU teaches “and obtaining, based on inference of the hyperparameter inference model, the hyperparameter of the operations research and optimization algorithm that corresponds to the feature of the data.”
See FU in [0137] describe " after obtaining a new set of hyperparameters, the values of the hyperparameters in the GAN are updated, and the next GAN training is started. This process is repeated n times, for example, 10 times, until the hyperparameters no longer change. At this point, the optimal GAN prediction model can be obtained."
Further, FU describes in [0139] "The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, and finally a prediction model is obtained to predict building energy consumption. " Here, FU mentions from [0137] and [0139] that after extracting the features, the features are then used as input for training the model. Then, the hyperparameters are optimized using the same model. FU also notes that a gradient enhancement algorithm to detect feature importance (i.e. an optimization algorithm that corresponds to the feature of the data) is also used.
Claim 3:
Regarding claim 3, FU in view of ACHIN, further in view of Gao teaches the limitations of claim 1. Referring to claim 3, FU further teaches “The method according to claim 1, wherein the method further comprises: analyzing the feature of the data;”
See FU in [0135] describe in step "S300: Use the XGBoost algorithm to perform feature selection on Rr;" Note the examiner construes analyzing to mean any type of evaluation of the information. FU here shows that feature selection counts as analyzing the feature of the data, where feature selection includes studying feature relevance, relationships, and significance to the target variable to choose the best subset.
Further, ACHIN teaches “and determining that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data, a user weight preference parameter in the data, or a problem structure parameter of the data,”
See ACHIN in paragraph [0224], mention "For the basic user, the predictive modeling system 100 may build the model immediately after the data set is defined, and the user interface 120 may include, but is not limited to, unanalysable data, To flag annoying issues that include too many results to expect good results, too many observations to perform in a reasonable amount of time, too many missing values, or variables whose distribution can lead to abnormal results." Further, see ACHIN in paragraph [0005] note “ The remaining variables that can be used to make the prediction may be referred to as "features," "predictors," or "independent variables." ACHIN in [0005] clarifies that variables can also mean features. In [0224], ACHIN describes that the variables whose distribution can lead to abnormal results correspond to (i.e. determining that the data of the current application scenario is abnormal data, wherein the feature of the data comprises at least one of distribution of the data). Refer to ACHIN in [0008] for further details on how different predictive modeling algorithms can be applied across different applications.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research. One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
Claim 8:
Regarding claim 8, FU in view of ACHIN, further in view of Gao, teach the limitation in claim 1.
Further, FU teaches “The method according to claim 1, wherein the obtaining data of a current application scenario and a feature of the data comprises: … and performing feature extraction on the data of the current application scenario to obtain the feature of the data,”
See FU describe in [0139] "the above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, and finally a prediction model is obtained to predict building energy consumption. " Here, FU shows extracting features from the energy consumption data to get these features used as input (i.e. obtain feature of the data) to the model for training.
Further, ACHIN teaches “obtaining the data of the current application scenario that is uploaded by the user through the user interface;”
See ACHIN in paragraph [0052] describe “In some embodiments, the actions of the method further identify, via the graphical user interface, the first and second features of the initial data set and model-specific predictions of the first and second features. Displaying graphical content. In some embodiments, the modeling procedure is a first modeling procedure that includes a particular modeling procedure associated with a particular prediction model," Later, see ACHIN mention in [0223] “To support the breadth of user requirements, some embodiments of the user interface 120 provide a set of interface tools that reflect the model building process. Each tool may also provide a series of features from basic to advanced. The first step in the model building process may involve loading and preparing a dataset. As discussed previously, a user may upload files or define how to access data from online systems. In the context of a modeling project group or hierarchy, the user may also specify which part of the parent dataset is to be used for the current project and which parts will be added." Note, examiner construes current application scenario to be a specific context or situation within which a machine learning model or system is applied to solve a particular problem or make predictions. Here, ACHIN mentions the user may determine the dataset to use for a current project, which in this case is a current application scenario for which a user uploads to run models to solve problems or make predictions. Here in [0223], ACHIN mentions a user may upload the data and build models for analysis using user interface 120 (i.e. obtain data of current application scenario uploaded by the user through the user interface).
Further, see ACHIN in [0103] describe “Accordingly, the user interface may be used by an analyst to enhance its own productivity and / or to improve the performance of the search engine 110. In some embodiments, the user interface 120 presents the results of the search in real time and allows the user to adjust the scope of the search in real time (e.g., to adjust the allocation of resources during evaluation of different modeling solutions). Can guide search. In some embodiments, the user interface 120 provides tools for coordinating the efforts of multiple data analysts working on the same prediction problem and / or related prediction problems.” Here, ACHIN shows that the user can use the user interface and tune the system according to the user’s preferences.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research with a user interface.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
Claim 9:
Regarding claim 9, FU in view of ACHIN, further in view of Gao teaches the limitations of claim 1.
Further, FU teaches “The method according to claim 1, wherein the obtaining data of a current application scenario and a feature of the data comprises: … and performing feature extraction on the data of the current application scenario to obtain the feature of the data.”
See FU in [0013] describe "The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, " Here, FU describes extracting features from data.
Further, ACHIN teaches “obtaining the data of the current application scenario that is uploaded by the user through the application programming interface;”
See ACHIN in [0304] describe “this conceptual model of information flow provides a context for the arrangement of functional modules in a layer. They are not simply stateless blocks that provide an application programming interface (API) to higher level blocks and consume APIs from lower level blocks. Rather, they are dynamic participants in collaboration and computation between users. FIG. 8 presents an arrangement of these functional modules. From the user's point of view, the interface service layer offers several distinct areas of functionality.” Here, ACHIN shows that API stands for an application programming interface.
Further, see ACHIN in [0235] describe "Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return." Here, ACHIN describes that user may upload data through an application programming interface to perform prediction. Further, see ACHIN in [0223] mentions “a user may upload files or define how to access data from online systems. In the context of a modeling project group or hierarchy, the user may also specify which part of the parent dataset is to be used for the current project and which parts will be added.” Note, examiner construes current application scenario to be a specific context or situation within which a machine learning model or system is applied to solve a particular problem or make predictions. Here, ACHIN mentions the user may determine the dataset to use for a current project, which in this case is a current application scenario for which a model is applied to solve problems or make predictions.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research with an application programming interface.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
Claim 11:
Regarding claim 11, FU teaches “to: obtain data of a current application scenario and a feature of the data;”
See FU in [0013] describe “The above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms.” Here, FU discloses obtaining energy consumption data (i.e. obtain data of a current application scenario) and extract high-level features (i.e. and a feature of the data).
Further, FU teaches “obtain a hyperparameter of an operations research and optimization algorithm based on the feature of the data and a hyperparameter inference model;”
See FU describe in [0137] "that after obtaining a new set of hyperparameters, the values of the hyperparameters in the GAN are updated, and the next GAN training is started. This process is repeated n times, for example, 10 times, until the hyperparameters no longer change. At this point, the optimal GAN prediction model can be obtained.”
Further, FU describes in [0139] "the above method collects building energy consumption data and related energy consumption feature data, uses variational autoencoder to extract high-level features, and uses extreme gradient enhancement algorithm to detect feature importance. Then, all the obtained energy consumption features are input into the generative adversarial network model for training. At the same time, the hyperparameters in the model are optimized using reinforcement learning algorithms, and finally a prediction model is obtained to predict building energy consumption. " This shows that FU teaches obtaining a new set of hyperparameters in [0137] that uses a reinforcement learning algorithm to help optimize the generative adversarial network model (i.e. hyperparameter inference model), and in [0139] all the features for training are then used as input for model training, then the hyperparameters are also optimized for the same model (i.e. obtain a hyperparameter of an operations research and optimization algorithm based on feature of data).
Further, see FU in [0107] mention “the server stores at least one executable instruction that enables the server to perform operations corresponding to the building energy consumption prediction method”, where FU shows this performs operations that relates to building a prediction method for energy consumption (i.e. part of an operations research and optimization algorithm). From the specification in paragraph [0003], which states that “an operations research and optimization algorithm is widely used in daily life and production practice... the operations research and optimization algorithm is indispensable and irreplaceable in many industries and fields such as social networking, entertainment, education, transportation, security, industry, logistics, and e-commerce.” This states that an operations research and optimization algorithm is involved in applications to various fields and industries, such as the example FU notes in building a method for predicting energy consumption.
However, FU did not teach “A computing device, comprising at least one processor and one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor,” Or “wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface;” or “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario;” or “and send a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm.”
In analogous system, ACHIN teaches “A computing device, comprising at least one processor and one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor,”
See ACHIN in [0084] describe “another embodiment of this aspect is a memory configured to store a machine-executable module that encodes a secondary prediction modeling procedure associated with the secondary prediction model, the memory comprising a secondary prediction modeling procedure. Is a memory including a plurality of tasks, including at least one pre-processing task and at least one model fitting task, a memory, and at least one processor configured to execute a machine-executable module, Executing the machine-executable module includes a predictive modeling device that includes a processor that causes the device to perform a secondary predictive modeling procedure on the fitted primary predictive model.” Here, ACHIN teaches a processor and a memory that can store instructions.
Further, ACHIN teaches “wherein the data of the current application scenario is uploaded by a user through a user interface or an application programming interface;”
See ACHIN in [0235] describe “Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return.” Here, ACHIN shows that for each data file, the user can upload this directly to an API or user interface.
Further, ACHIN teaches “and send a result of the current application scenario to the user through the user interface or the application programming interface, wherein the result of the current application scenario is obtained bases on the data of the current application scenario, the hyperparameter, and the operations research and optimization algorithm.”
See ACHIN in [0235] describe “Service-based prediction may occur either interactively or via an API. For bi-directional prediction, the user may enter feature values for each new observation, or upload a file containing data for one or more observations. The user may then receive the predictions directly through the user interface 120 or download them as a file. For API predictions, the external system may access the prediction module via a local or remote API, submit one or more observations, and receive corresponding calculated predictions in return.” Here, ACHIN shows that for each data file, the user can upload this directly to an API or user interface. The user also receives calculated predictions or results from the API interface.
Further, see ACHIN describe in [0293] that “(6) External system 660. As with any other Internet application, the use of APIs may allow external systems to integrate with the predictive modeling system 100 at any layer of the architecture 600. For example, a business dashboard application can access graphical visualization and modeling results through the interface services layer 620. An external data warehouse or even a live business application can provide the modeled dataset to the analytics service layer 630 through a data integration platform. The reporting application can access all modeling results from a particular time period through the data services layer 650.” Here, ACHIN shows the results from a live business application (i.e. current application scenario) are sent through the API, and from [0235] the results from the API can be sent to a user.
Further, see ACHIN in [0161] mention “Returning to FIG. 3, at step 340 of method 300, a result of executing the selected modeling procedure according to the resource allocation schedule may be received. These results may include one or more predictive models generated by the executed modeling procedure. In some embodiments, the execution of the modeling procedure may include fitting the prediction model to one or more datasets associated with the prediction problem, so that the prediction model received in step 340 may include the prediction problem Fitted to the dataset associated with. Fitting the predictive model to the data set of the predictive problem includes adjusting one or more hyperparameters of a predictive modeling procedure that generates the predictive model, adjusting one or more parameters of the generated predictive model.” Here, ACHIN shows the information that the current application scenario of ‘fitting the prediction model to one or more datasets associated with the prediction problem’, is based on adjusting one or more hyperparameters (i.e. hyperparameter) of a predictive modeling procedure (i.e. operations research and optimization algorithm).
Later, see ACHIN mention in [0223] “To support the breadth of user requirements, some embodiments of the user interface 120 provide a set of interface tools that reflect the model building process. Each tool may also provide a series of features from basic to advanced. The first step in the model building process may involve loading and preparing a dataset. As discussed previously, a user may upload files or define how to access data from online systems. In the context of a modeling project group or hierarchy, the user may also specify which part of the parent dataset is to be used for the current project and which parts will be added." Note, examiner construes current application scenario to mean a specific context or situation within which a machine learning model or system is applied to solve a particular problem or make predictions. Here, ACHIN mentions the user may determine the dataset to use for a current project, which in this case is a current application scenario for which a user uploads to run models to solve problems or make predictions. See ACHIN in [0242] describe for more details. Here, ACHIN shows that the results of a prediction modeling system 100 are stored until needed.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method along with a processor and a memory for optimizing a model for operations research with analyzing a calculated result.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
However, FU in view of ACHIN did not teach “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario;”
In an analogous field, Gao teaches “wherein the hyperparameter inference model is obtained through dynamic training based on training data obtained in a historical application scenario and training data obtained in the current application scenario;”
See Gao in [0070] describe "All these experiments show that the proposed novel class incremental learning method, SupportNet, solves the catastrophic forgetting problem by combining the strength of deep learning and SVM. SupportNet can efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. With the help of two powerful consolidation regularizers, the support data can effectively help the deep learning model prevent the catastrophic forgetting issue, eliminate the necessity of retraining the model from scratch, and maintain a stable learned representation that corresponds to the old and the new data." Here, Gao shows that the model called SupportNet can ‘efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information’ to indicate training uses both old data (i.e. data from historical application scenario) and new data (i.e. data obtained in the current application scenario).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the references of FU and ACHIN, with the teachings of Gao by using the teachings of FU and ACHIN of optimizing a model with feature analysis, with Gao’s teaching of dynamic training using historical data and current application data.
One of ordinary skill in the art would be motivated to do so because by integrating Gao’s framework into the methods of FU and ACHIN, one with ordinary skill in the art would achieve a method like “SupportNet can efficiently identify the important information associated with the old data, which is fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. With the help of two powerful consolidation regularizers, the support data can effectively help the deep learning model prevent the catastrophic forgetting issue, eliminate the necessity of retraining the model from scratch, and maintain a stable learned representation that corresponds to the old and the new data,” (see Gao in [0070]), and “the present method is a highly effective framework (called SupportNet in the following), which can perform class incremental learning without catastrophic forgetting,” (see Gao in [0055]).
Claim 12:
Regarding claim 12, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 11. Regarding claim 12, the claim recites similar limitations as corresponding claim 2 and is rejected for similar reasons as claim 2 using similar teachings and rationale.
Claim 13:
Regarding claim 13, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 11. Regarding claim 13, the claim recites similar limitations as corresponding claim 3 and is rejected for similar reasons as claim 3 using similar teachings and rationale.
Claim 18:
Regarding claim 18, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 11. Regarding claim 18, the claim recites similar limitations as corresponding claim 8 and is rejected for similar reasons as claim 8 using similar teachings and rationale.
Claim 19:
Regarding claim 19, FU in view of ACHIN, further in view of Gao teach the limitations in claim 11. Regarding claim 19, the claim recites similar limitations as corresponding claim 9 and is rejected for similar reasons as claim 9 using similar teachings and rationale.
Claims 4 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over FU in view of ACHIN, further in view of Gao, and further in view of Ruan, H. et al., (Pub. No. CN107528722A) , published on December 29, 2017, (hereafter, RUAN).
Claim 4:
Regarding claim 4, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 1.
Further, ACHIN teaches “The method according to claim 1, wherein the method further comprises: analyzing the result of the current application scenario;”
See ACHIN in [0077] describe “performing the nested cross-validation further comprises: testing a first fitted quadratic model and a second fitted quadratic model in a second partition of the data set; Comparing the first fitted quadratic model to the second fitted quadratic model based on the results of testing the first and second fitted quadratic models in the second partition.” Here, ACHIN describes the analysis step as a comparison of a first fit model to a second fit model.
Later, see ACHIN mention in [0223] “To support the breadth of user requirements, some embodiments of the user interface 120 provide a set of interface tools that reflect the model building process... The first step in the model building process may involve loading and preparing a dataset… In the context of a modeling project group or hierarchy, the user may also specify which part of the parent dataset is to be used for the current project and which parts will be added." Note, examiner construes current application scenario to mean a specific context or situation within which a machine learning model or system is applied to solve a particular problem or make predictions. Here, ACHIN mentions the user may determine the dataset to use for a current project, which in this case is a current application scenario for which a user uploads to run models to solve problems or make predictions.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research with analyzing a calculated result.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
However, FU in view of ACHIN, further in view of Gao, did not teach “and when the result of the current application scenario does not meet a preset condition, determining that the data of the current application scenario is abnormal data,”
In an analogous field, RUAN teaches “and when the result of the current application scenario does not meet a preset condition, determining that the data of the current application scenario is abnormal data,”
See RUAN in [0063-0067] describe “Specifically, as shown in Figure 4, the clustering prediction of the time subsequence containing the discrete points to determine the first prediction result of the discrete points may include: S401: Obtain the time subsequence within a preset window containing the discrete points; S403: Calculate the distance between the time subsequence distance and the cluster centers of multiple preset categories, wherein the multiple preset categories include categories obtained based on historical time series clustering analysis of the business indicators; S405: Determine whether the minimum distance among the distances is greater than a preset distance threshold; S407: If the judgment result is yes, then the discrete point is determined to be an outlier.” Here, RUAN shows determining if result of a calculation of minimum range is more than a pre-determined threshold (i.e. not meet a preset condition), determine that point as an abnormal data point. Since this method applies to multiple data points, this method will evaluate any data points that can be abnormal data. See RUAN in [0068] for more information.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the references of FU, ACHIN, and Gao with the teachings of RUAN by using the teachings of FU, ACHIN, and Gao of optimizing a model with feature analysis, with RUAN’s teaching of when the calculation result does not meet a preset condition, determining that the data of the current application scenario is abnormal data.
One of ordinary skill in the art would be motivated to do so because by integrating RUAN’s framework into the methods of FU, ACHIN, and Gao, one with ordinary skill in the art would achieve the goal of providing a method where “the advantage of EGADS is that its anomaly detection module integrates multiple different anomaly detection algorithms,” ([0039], RUAN ).
Claim 14:
Regarding claim 14, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 11. Regarding claim 14, the claim recites similar limitations as corresponding claim 4 and is rejected for similar reasons as claim 4 using similar teachings and rationale.
Claims 5 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over FU in view of ACHIN, further in view of Gao, and further in view of Feinman, R. et al., (Patent Publication No. US10572823B1), published on February 25, 2020, (hereafter, FEINMAN).
Claim 5:
Regarding claim 5, FU in view of ACHIN, further in view of Gao, teach the limitations in claim 3. However, FU in view of ACHIN, further in view of Gao did not teach “the method according to claim 3, wherein the method further comprises: optimizing the hyperparameter of the operations research and optimization algorithm by using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result,”
In an analogous art, FEINMAN teaches “the method according to claim 3, wherein the method further comprises: optimizing the hyperparameter of the operations research and optimization algorithm by using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result,”
See FEINMAN in col. 8, lines 20-36 describe "Optimization module 106 may identify the set of hyperparameters that optimizes objective function 130 in a variety of ways. In one example, optimization module 106 may calculate a value of objective function 130 for each candidate hyperparameter set within candidate hyperparameter sets 124. In this example, optimization module 106 may determine that optimal hyperparameter set 126 generated the highest value, relative to the values generated by the other candidate hyperparameter sets within candidate hyperparameter sets 124, and select optimal hyperparameter set 126 based on the determination. Optimization module 106 may calculate the value of objective function 130 for each candidate hyperparameter set in various ways. For example, optimization module 106 may calculate the value of objective function 130 for each candidate hyperparameter set based on how each candidate hyperparameter set performs in a testing environment." Here, FEINMAN shows that the optimization module 106 is responsible in identifying an optimized hyperparameter set (i.e. to obtain an optimized hyperparameter), and the highest value of objective function (i.e. an optimized calculation result).
See FEINMAN in col. 6, lines 32-40 mention that "the steps shown in FIGS. 3-4 may be performed by any suitable computer-executable code and/or computing system, including system 100 in FIG. 1, system 200 in FIG. 2, and/or variations or combinations of one or more of the same. In one example, each of the steps shown in FIGS. 3-4 may represent an algorithm whose structure includes and/or is represented by multiple sub-steps, examples of which will be provided in greater detail below." Here, FEINMAN also notes that the optimization process also includes an algorithm (i.e. using a hyperparameter optimization algorithm).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of FU, ACHIN, and Gao with the teachings of FEINMAN by using the teachings of FU, ACHIN, and Gao of optimizing a model with feature analysis, with FEINMAN’s teaching of using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result.
One of ordinary skill in the art would be motivated to do so because by integrating FEINMAN’s framework into the methods of FU, ACHIN, and Gao, one with ordinary skill in the art would achieve the goal of providing a method that would “be able to (i) improve the deployability of malware detection models and (ii) improve the classification accuracy of malware detection models, thereby reducing the number of resulting false positives and/or false negatives, when compared to traditional malware detection models created with hyperparameters optimized for efficacy ” (See FEINMAN in col. 3, line 67 through col. 4, lines 1-6).
Claim 15:
Regarding claim 15, FU in view of ACHIN, further in view of Gao, teaches the limitations in claim 13. Regarding claim 15, the claim recites similar limitations as corresponding claim 5 and is rejected for similar reasons as claim 5 using similar teachings and rationale.
Claims 6, 7, 16, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over FU in view of ACHIN, further in view of Gao, further in view of FEINMAN, and further in view of Dupont, L. et al. (US PG Pub. No. US20120137367 A1), published on May 31, 2012, (hereafter, DUPONT).
Claim 6:
Regarding claim 6, FU in view of ACHIN, further in view of Gao, further in view of FEINMAN teaches the limitations in claim 5.
However, FU in view of ACHIN, further in view of Gao, further in view of FEINMAN did not teach “the method according to claim 5, wherein the method further comprises: recording the abnormal data and the optimized calculation result corresponding to the abnormal data into a training data set used to train the hyperparameter inference model,”
In an analogous field, DUPONT teaches “the method according to claim 5, wherein the method further comprises: recording the abnormal data and the optimized calculation result corresponding to the abnormal data into a training data set used to train the hyperparameter inference model,”
See DUPONT in [0628-0629] describe “alternatively, compute the normalized log-probability of the input workflow instance with respect to the workflow model. If this probability is within a given multiple of standard deviations (typically 3) from the average probability of workflow instances [134] within the training set, this instance [134] is flagged as an outlier. It should be noted that outliers, as well as anomalies, are defined by assessing the normalcy of an input workflow instance [134] with respect to a model derived from the observation of instances [134] as training set. As is the case for the general case of detecting anomalies by deviation [805], this training set or referential can be defined in a number of ways.” Here, DUPONT notes that outlier (i.e. abnormal data) and the normalized log-probability of the input workflow instance (i.e. optimized calculation result corresponding to the abnormal data) are part of the training data set that is used to train the model. The optimization term is construed by examiner to mean any type of optimizing to improve model performance. For example, here, DUPONT mentions that normalizing the log probability means taking the logarithm of a probability value that has been scaled to ensure a set of probabilities sums to 1. This helps transform small probabilities into manageable, negative real numbers, improving numerical stability and computational speed and is part of optimized result that relates to the outliers. Further, DUPONT mentions in [1086] that “Logging [313]: when an anomaly is detected, the logging functionality either stores metadata relative to the data associated to this anomaly, or stores the native data itself. In the latter case, the way data items are stored can be configured more finely” Here, DUPONT shows recording the data that includes both the anomaly or abnormal data, as well as any of the data that describes the anomaly including the calculations of normalized log-probability described in [0628].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of FU, ACHIN, Gao, and FEINMAN, with the teachings of DUPONT by using the teachings of FU, ACHIN, Gao, and FEINMAN, of optimizing a model with feature analysis, with DUPONT’s teaching of using recorded abnormal data and an optimized calculation result corresponding to the abnormal data for training an inference model.
One of ordinary skill in the art would be motivated to do so because by integrating DUPONT’s framework into the methods of FU, ACHIN, Gao, and FEINMAN, one with ordinary skill in the art would achieve the goal of providing a method that “efficiently performs continuous monitoring of data produced or circulating within an entity, or social network, … and uses novel behavioral analysis techniques in order to detect, report and/or make users aware of possibly preventable damaging events, of an accidental or fraudulent nature,” (DUPONT, [0150]).
Claim 7:
Regarding claim 7, FU in view of ACHIN, further in view of Gao, in view of FEINMAN, further in view of DUPONT, teaches the limitations of claim 6.
Further, FU teaches “The method according to claim 6, wherein the method further comprises: determining that the hyperparameter inference model is to be updated;”
See FU in [0194] describe “In this embodiment, after completing one training of the GAN prediction model as described in step S400 above, the hyperparameters in GAN, LSTM and CNN are optimized using reinforcement learning algorithms to find the best combination of hyperparameters and update them. Then, the next training of the GAN prediction model is carried out until the optimal GAN prediction model is obtained.” Here, FU shows that model is updated when it is optimized by algorithms to find the best combination of hyperparameters until an optimal prediction model is obtained. See FU in [0136] for details.
Further, see FU in [0137] describe "It should be noted that after obtaining a new set of hyperparameters, the values of the hyperparameters in the GAN are updated, and the next GAN training is started. This process is repeated n times, for example, 10 times, until the hyperparameters no longer change. At this point, the optimal GAN prediction model can be obtained. " FU here teaches the optimal GAN prediction model relates to an updated hyperparameter inference model. See FU in [0194] and [0212-0213] for more information.
Further, ACHIN teaches “and training the hyperparameter inference models, based on training data in the training data set, to obtain an updated hyperparameter inference model.”
See ACHIN in [0415] describe “include real-world data combined with machine-generated data (for the purpose of covering gender) or data completely generated by machine-based probabilistic models. In some embodiments, the value of the target variable used to train the secondary model is a predicted value from the primary model.” Further, see ACHIN in [0265] describe “the selection of the final model can be made by the predictive modeling system 100 or by the user. In the latter case, the predictive modeling system may allow the user to, for example, evaluate the model's ranked verification set performance, compare the performance and rank by quality measures other than those used in the fitting process, and Support may be provided to assist in making this determination, including the opportunity to build an ensemble model from these component models that exhibit the best individual performance.” Here, ACHIN describes in [0415] training more than one model and in [0265] the method selects the best performing model (i.e. updated model). See ACHIN in [0461-0462] for details.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research with training multiple models to obtain an updated model.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
Claim 16:
Regarding claim 16, FU in view of ACHIN, further in view of Gao, and further in view of FEINMAN, teaches the limitations of claim 15. Regarding claim 16, the claim recites similar limitations as corresponding claim 6 and is rejected for similar reasons as claim 6 using similar teachings and rationale.
Claim 17:
Regarding claim 17, FU in view of ACHIN, further in view of Gao, further in view of FEINMAN, and further in view of DUPONT, teaches the limitations of claim 16. Regarding claim 17, the claim recites similar limitations as corresponding claim 7 and is rejected for similar reasons as claim 7 using similar teachings and rationale.
Claims 10 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over FU in view of ACHIN, further in view of Gao, and further in view of Bruckhaus, T. et al., (US Patent No. US 8417715B1), published on April 9, 2013, (hereafter, BRUCKHAUS).
Claim 10:
Regarding claim 10, FU in view of ACHIN, further in view of Gao, teach the limitations of claim 1.
Further, ACHIN teaches “The method according to claim 1, wherein the method further comprises: obtaining an operations research and optimization task type configured by the user;”
See ACHIN in [0183] describe "Analysis of the data set may be performed using any suitable technique. The variable importance, which measures the degree of significance that each feature has in predicting the target, is “gradient boost tree”, “random forest” of Breiman and Cutler, “alternating conditional expectation”, and / or The analysis may be performed using other suitable techniques. Variable effects, which measure the direction and size of the effect of the feature on the target, may be analyzed using "normalized regression", "logistic regression", and / or other suitable techniques. Impact hotspots, which identify areas where features provide the most information in predicting a target, may be analyzed using a "RuleFit" algorithm and / or other suitable techniques." See ACHIN in [0401] for further details. Note: task type is construed by examiner to mean various tasks that machine learning models can perform, such as for classification or regression, etc. Here, ACHIN discusses various task types where analysis may be performed using other suitable techniques including regression.
Further, see ACHIN in [0181] "At step 406 of method 400, search engine 110 prompts the user to identify which of the variables are targets and / or which are features. In some embodiments, search engine 110 may also include a model performance metric used to score the model (e.g., a statistical optimization technique, such as a statistical learning algorithm implemented by search engine 110). In a sense, it prompts the user to identify the model performance metric to be optimized).” Here, ACHIN describes that a user can identify information such as variables, or performance metrics, and optimization type.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to combine the base reference of FU and incorporate into the teachings of ACHIN because both references teach a method for optimizing a model for operations research with analyzing obtaining an operations research and optimization task type configured by a user.
One of ordinary skill in the art would be motivated to do so because this would help “guide the allocation of resources to evaluating predictive modeling procedures, to feature engineering tasks, and to mixing predictive models, thereby providing potential predictive modeling for predictive problems. It can facilitate a cost-effective assessment of the space of the technique,” (ACHIN, [0062]).
However, FU in view of ACHIN, further in view of Gao, did not teach “and determining the operations research and optimization algorithm based on the task type.”
In an analogous art, BRUCKHAUS teaches “and determining the operations research and optimization algorithm based on the task type.”
See BRUCKHAUS in col. 1, lines 50-62, describe "the application of analytics to a business task or query involves collecting data about the problem; storing, managing, integrating, and preparing such data; and applying mathematical, quantitative and statistical analysis, and predictive modeling. As a result of applying analytics, organizations can better understand business needs and issues, discover causes and opportunities, predict risk levels and events, take steps to prevent risks and events, and perform other similar activities that are beneficial to the organization. It should be appreciated that analytics are used not only in the commercial for-profit sector but also in government, the non-profit sector, and by individuals to evaluate and solve many different types of problems or queries." Here, BRUCKHAUS describes that analytics including predictive modeling and statistical or math analysis can apply to different types of problems (i.e. task type).
Further, see BRUCKHAUS in col. 64, lines 35-45 mention “as described above, the invention selects algorithms, algorithm parameters and inputs to maximize the objective function, and the user can either specify the objective function directly, or the model manager 144 can discover the objective function using CQM and Optimization.” Here, BRUCKHAUS shows the user can specify an objective function and let the system select algorithms that maximize model performance. Also see BRUCKHAUS in figure 17, where an algorithm is selected from a library to improve the model performance of specific models.
PNG
media_image2.png
996
756
media_image2.png
Greyscale
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of FU, ACHIN, and Gao with the teachings of BRUCKHAUS by using the teachings of FU, ACHIN, and Gao of optimizing a model with feature analysis, with the teaching of BRUCKHAUS of using a hyperparameter optimization algorithm to obtain an optimized hyperparameter and an optimized calculation result.
One of ordinary skill in the art would be motivated to do so because by integrating BRUCKHAUS’s framework into the methods of FU and ACHIN, and Gao, one with ordinary skill in the art would achieve the goal of providing “a further advantage to be able to improve analytic models over time. That is, data mining could be applied to the data-mining models to improve the analytic models without a user being required to be an expert in applying complex data management or data mining techniques,” (See BRUCKHAUS in col. 3, lines 46-51).
Claim 20:
Regarding claim 20, FU in view of ACHIN, further in view of Gao, teaches the limitations of claim 11. Regarding claim 20, the claim recites similar limitations as corresponding claim 10 and is rejected for similar reasons as claim 10 using similar teachings and rationale.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WENWEI ZENG whose telephone number is (571)272-7111. The examiner can normally be reached Monday-Friday, 8am-5pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Usmaan Saeed can be reached at (571) 272-4046. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/WenWei Zeng/Examiner, Art Unit 2146
/USMAAN SAEED/Supervisory Patent Examiner, Art Unit 2146
Prosecution Insights