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 .
Response to Amendment
This action is in response to applicant’s arguments and amendments filed 6/10/2026, which are in response to USPTO Office Action mailed 3/19/2026. Applicant’s arguments have been considered with the results that follow: THIS ACTION IS MADE FINAL.
Status of Claims
Claims 16, 20 and 28-45 are currently pending in the present application. Claims 1-15, 17-19 and 21-27 are currently cancelled.
Claim Rejections - 35 USC § 103
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.
Claim(s) 16, 20, 28-32 and 36-41 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tong (US Patent No.: 11,429,893; Date of Patent; Aug. 30, 2022) in view of Ghassal et al. (US PGPUB No. 2024/0003250; Pub. Date; Jan. 4, 2024) and MANIVANNAN et al. (International Publication No.: WO 2023/223289; Pub. Date: Nov. 23, 2023).
Regarding independent claim 16,
Tong discloses a parallelized database system comprises: a plurality of computing device clusters, wherein a first computing device cluster of the plurality of computing device clusters includes a first plurality of computing devices of pluralities of computing devices, wherein a first computing device of the first plurality of computing devices includes a first plurality of computing nodes of pluralities of computing nodes, and wherein a first computing node of the first plurality of computing nodes includes a first plurality of processing core resources of pluralities of processing core resources, See FIG. 1 & Col. 4, lines 30-32, (Disclosing a system for massively-parallel real-time database-integrated machine learning inference. The system may deploy an ML model associated with a virtual table or function. A database instance includes a cluster of nodes which may be implemented by a virtual machine or physical host, i.e. a parallelized database system comprises: a plurality of computing device clusters. The cluster of nodes including a leader node 132A and one or more compute nodes 134A, i.e. wherein a first computing device cluster of the plurality of computing device clusters includes a first plurality of computing devices of pluralities of computing devices (e.g. the virtual machine/physical host device implementing a device cluster), wherein a first computing device of the first plurality of computing devices includes a first plurality of computing nodes of pluralities of computing nodes (e.g. leader node 132A and compute nodes 134A are a plurality of nodes).) See Col. 6, lines 20-23, (A database instance 130A is associated with computing resource capabilities including an amount of memory, processing capabilities, network bandwidth, number of virtual or physical machines implementing the database instance, etc. FIG. 1 illustrates a database instance 130A as comprising leader node 132A and compute nodes 134A, therefore the plurality of nodes are associated with the computing resources of database instance 130A, wherein a first computing node of the first plurality of computing nodes includes a first plurality of processing core resources of pluralities of processing core resources.)
determine coefficients of the function; See Col. 14, lines 47-61, (A virtual machine instance 622 may train a machine learning model by identifying values for certain parameters including coefficients, weights, centroids, etc., i.e. determine coefficients of the function;)
generate a query plan regarding the model training operation, wherein the query plan identifies a second plurality of processing core resources of the pluralities of processing core resources and identifies a plurality of machine learning training data corresponding to the dataset; See FIG. 5, (FIG. 5 illustrates method 500 comprising step505 of obtaining an ML model. As noted in Col. 5, lines 20-24, ML models may be configured based on user inputs providing ML configuration metadata, i.e. generate a query plan regarding the model training operation (e.g. providing an ML model includes the process wherein a user provides metadata for deploying an ML model which comprises the series of steps illustrated in FIG. 5). Method 500 then comprises step 520 of causing an endpoint to be configured within the provider network to be associated with one or more model serving units, i.e. wherein the query plan identifies a second plurality of processing core resources of the pluralities of processing core resources (e.g. the endpoints within the provider network).) See Col. 10, lines 31-40, (Step 505 includes receiving the ML model from a machine learning service that trained the model. Note Col. 14, lines 26-31 wherein model training system 620 retrieves training data from a training data store 660 for training an ML model to be deployed, i.e. identifies a plurality of machine learning training data corresponding to the dataset;)
execute by the second set of processing core resources, the function on the plurality of machine learning training data to produce a plurality of values for the coefficients; See Col. 19, lines 60-65, (A deployment request comprises an identifier of an endpoint and an identification of trained machine learning models to be deployed.) See Col. 21, lines 31-44, (A virtual machine instance 642 may execute code 656 stored in an identified ML scoring container 650 in response to an execution request. Executable instructions in code 656 read a model data file to determine values of the coefficients, etc. associated with the requested ML model, i.e. execute by the second set of processing core resources (e.g. the ML model is deployed to a requesting endpoint), the function on the plurality of machine learning training data to produce a plurality of values for the coefficients (e.g. executing the ML model request includes executing code that determines model data including coefficients);)
and store the function and the desired set of values for the coefficients as the machine learning model. See Col. 5, lines 7-14, (Storage service 108 may store a trained ML model that may be retrieved by a management module 116. Note Col. 15, lines 10-15 wherein model data includes characteristics of the machine learning model being trained, such as a number of layers in the machine learning model, hyperparameters of the machine learning model, coefficients of the machine learning model, weights of the machine learning model, etc., i.e. store the function and the desired set of values for the coefficients as the machine learning model.
Tong does not disclose the step of process[ing] the plurality of values for the coefficients to identify a desired set of values for the coefficients;
Ghassal discloses the step of process[ing] the plurality of values for the coefficients to identify a desired set of values for the coefficients; See FIG. 4A & Paragraph [0049], (FIG. 4 illustrates method 400 comprising step 408 of training a machine learning model to make pore pressure predictions using offset well data. Training of the model may involve adjusting model parameters including weight coefficients to their optimal values, i.e. process the plurality of values for the coefficients to identify a desired set of values for the coefficients (e.g. the optimal coefficients).)
Tong and Ghassal are analogous art because they are in the same field of endeavor, machine learning model configuration. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of Tong to include the method of optimizing coefficients for an ML model as disclosed by Ghassal. Paragraph [0022] of Ghassal discloses that the process includes updating prognostic conditions in real-time using input data acquired during a drilling process. This represents an improvement in accuracy and reduction in uncertainty of a prediction that is driven by geological data and nonlinear approximation capability of machine learning.
The examiner notes that while Ghassal is directed to facilitating drilling operations, one of ordinary skill in the art would be able to recognize that a machine learning model may be trained on any desired data and is not merely limited in scope to drilling.
Tong-Ghassal does not disclose the step wherein a set of processing core resources of the pluralities of processing core resources is operable to: obtain within a query regarding a dataset, a model training operation to generate and initiate training of a machine learning model;
in response to the training model operation: determine a function of the machine learning model;
MANIVANNAN discloses the step wherein a set of processing core resources of the pluralities of processing core resources is operable to: obtain within a query regarding a dataset, a model training operation to generate and initiate training of a machine learning model; See Pg. 2, Paragraph 5, (Disclosing a method for generating and training an ML model. Initially, a user 102 may submit a request to generate an ML model. The system may generate the ML model based on a software development kit (SDK) and launch instance function parameters on the selected functionality. The ML model generating system 101 determines at least one of a run identifier, accuracy score and hyper-setting parameter for the generated ML model. When the run identifier, accuracy score and hyper-setting parameter are below a threshold limit, the ML model generating system 101 trains the generated ML model using launch instance function parameters.
in response to the training model operation: determine a function of the machine learning model; See Pg. 2, Paragraph 5, (After training, when the run identifier, accuracy score and hyper setting parameter are above a pre-defined threshold limit, the system may store the ML model in the database.) See Pg. 7, Paragraph 12, (Generating module 204 is configured to train the generated ML model using launch instance function parameters when the at least one run identifier, accuracy score and hyper setting parameter is below a threshold limit, i.e. in response to the training model operation: determine a function of the machine learning model (e.g. the run identifier, accuracy score and hyper setting parameters of the generated model inform its functionality.).)
Tong-Ghassal and MANIVANNAN are analogous art because they are in the same field of endeavor, machine learning model generation. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of Tong-Ghassal to include the method of generating and training an ML model for deployment in response to a user request as disclosed by MANIVANNAN. Pg. 12, paragraphs 6-11 of MANIVANNAN disclose that the ML model generating system helps a user select a best-fit ML model by providing suggestions in terms of functionalities, SDK and launch instance function parameters in order to facilitate obtaining accurate results. The ML model generation system is implemented as a cloud server which optimizes infrastructure, enhances security and increases connectivity/accessibility in terms of user outreach.
Regarding independent claim 20,
MANIVANNAN discloses the step a non-transitory computer readable storage medium comprises: a first memory section that stores operational instructions that, when executed by a set of processing core resources of pluralities of processing core resources of a parallelized database system, causes the set of processing core resources to: obtain within a query regarding a dataset, a model training operation to generate and initiate training of a machine learning model; See Pg. 2, Paragraph 5, (Disclosing a method for generating and training an ML model. Initially, a user 102 may submit a request to generate an ML model. The system may generate the ML model based on a software development kit (SDK) and launch instance function parameters on the selected functionality. The ML model generating system 101 determines at least one of a run identifier, accuracy score and hyper-setting parameter for the generated ML model. When the run identifier, accuracy score and hyper-setting parameter are below a threshold limit, the ML model generating system 101 trains the generated ML model using launch instance function parameters, i.e. non-transitory computer readable storage medium comprises: a first memory section that stores operational instructions that (e.g. Note Pg. 12, paragraph 5 wherein the system may be implemented as a computer-readable storage medium storing instructions for execution by one or more processors), when executed by a set of processing core resources of pluralities of processing core resources of a parallelized database system (e.g. Note Pg. 13, paragraph 9 wherein the process of generating an ML model may be performed in parallel by multiple distributed processing units), causes the set of processing core resources to: obtain within a query regarding a dataset (e.g. Note Pg. 3, paragraph 2 wherein launch instance function parameters 209 may include one or more datasets), a model training operation to generate and initiate training of a machine learning model (e.g. an ML model is generated in response to a user request);)
in response to the training model operation: determine a function of the machine learning model; See Pg. 2, Paragraph 5, (After training, when the run identifier, accuracy score and hyper setting parameter are above a pre-defined threshold limit, the system may store the ML model in the database.) See Pg. 7, Paragraph 12, (Generating module 204 is configured to train the generated ML model using launch instance function parameters when the at least one run identifier, accuracy score and hyper setting parameter is below a threshold limit, i.e. in response to the training model operation: determine a function of the machine learning model (e.g. the run identifier, accuracy score and hyper setting parameters of the generated model inform its functionality.).)
MANIVANNAN does not disclose the step wherein the system may determine coefficients of the function;
generate a query plan regarding the model training operation, wherein the query plan identifies a second set of processing core resources of the pluralities of processing core resources and identifies a plurality of machine learning data corresponding to the dataset;
and a second memory section that stores operational instructions that, when executed by the of second set of processing core resources, causes the second set of processing core resources to: execute by the second set of processing core resources, the function on the plurality of machine learning training data to produce a plurality of values for the coefficients;
and store the function and the desired set of values for the coefficients as the machine learning model.
Tong discloses a system configured to determine coefficients of the function; See Col. 14, lines 47-61, (A virtual machine instance 622 may train a machine learning model by identifying values for certain parameters including coefficients, weights, centroids, etc., i.e. determine coefficients of the function;)
generate a query plan regarding the model training operation, wherein the query plan identifies a second set of processing core resources of the pluralities of processing core resources and identifies a plurality of machine learning data corresponding to the dataset; See FIG. 5, (FIG. 5 illustrates method 500 comprising step505 of obtaining an ML model. As noted in Col. 5, lines 20-24, ML models may be configured based on user inputs providing ML configuration metadata, i.e. generate a query plan regarding the model training operation (e.g. providing an ML model includes the process wherein a user provides metadata for deploying an ML model which comprises the series of steps illustrated in FIG. 5). Method 500 then comprises step 520 of causing an endpoint to be configured within the provider network to be associated with one or more model serving units, i.e. wherein the query plan identifies a second set of processing core resources of the pluralities of processing core resources (e.g. the endpoints within the provider network).) See Col. 10, lines 31-40, (Step 505 includes receiving the ML model from a machine learning service that trained the model. Note Col. 14, lines 26-31 wherein model training system 620 retrieves training data from a training data store 660 for training an ML model to be deployed, i.e. identifies a plurality of machine learning training data corresponding to the dataset;)
and a second memory section that stores operational instructions that, when executed by the of second set of processing core resources, causes the second set of processing core resources to: execute by the second set of processing core resources, the function on the plurality of machine learning training data to produce a plurality of values for the coefficients; See Col. 19, lines 60-65, (A deployment request comprises an identifier of an endpoint and an identification of trained machine learning models to be deployed.) See Col. 21, lines 31-44, (A virtual machine instance 642 may execute code 656 stored in an identified ML scoring container 650 in response to an execution request. Executable instructions in code 656 read a model data file to determine values of the coefficients, etc. associated with the requested ML model, i.e. a second memory section that stores operational instructions that, when executed by the of second set of processing core resources (e.g. an identified endpoint comprises a computing system/device configured to execute an ML model), causes the second set of processing core resources to: execute by the second set of processing core resources, the function on the plurality of machine learning training data to produce a plurality of values for the coefficients (e.g. executing the ML model request includes executing code that determines model data including coefficients);)
and store the function and the desired set of values for the coefficients as the machine learning model. See Col. 5, lines 7-14, (Storage service 108 may store a trained ML model that may be retrieved by a management module 116. Note Col. 15, lines 10-15 wherein model data includes characteristics of the machine learning model being trained, such as a number of layers in the machine learning model, hyperparameters of the machine learning model, coefficients of the machine learning model, weights of the machine learning model, etc., i.e. store the function and the desired set of values for the coefficients as the machine learning model.
MANIVANNAN and Tong are analogous art because they are in the same field of endeavor, machine learning model generation. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of MANIVANNAN to include the method of assessing coefficients for generating ML models as disclosed by Tong. Col. 3, lines 1-20 of Tong disclose that the system may use a massively parallel real-time inference database engine that can be integrated with cloud-based databases to allow users to use standard query interfaces to query ML models to answer predictive-type questions. This allows users to query a data warehouse or data lake through a database engine and perform inferential analysis through existing tools.
MANIVANNAN-Tong does not disclose the step of process[ing] the plurality of values for the coefficients to identify a desired set of values for the coefficients;
Ghassal discloses the step of process[ing] the plurality of values for the coefficients to identify a desired set of values for the coefficients; See FIG. 4A & Paragraph [0049], (FIG. 4 illustrates method 400 comprising step 408 of training a machine learning model to make pore pressure predictions using offset well data. Training of the model may involve adjusting model parameters including weight coefficients to their optimal values, i.e. process the plurality of values for the coefficients to identify a desired set of values for the coefficients (e.g. the optimal coefficients).)
MANIVANNAN, Tong and Ghassal are analogous art because they are in the same field of endeavor, machine learning model configuration. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of MANIVANNAN-Tong to include the method of optimizing coefficients for an ML model as disclosed by Ghassal. Paragraph [0022] of Ghassal discloses that the process includes updating prognostic conditions in real-time using input data acquired during a drilling process. This represents an improvement in accuracy and reduction in uncertainty of a prediction that is driven by geological data and nonlinear approximation capability of machine learning.
The examiner notes that while Ghassal is directed to facilitating drilling operations, one of ordinary skill in the art would be able to recognize that a machine learning model may be trained on any desired data and is not merely limited in scope to drilling.
Regarding dependent claim 28,
As discussed above with claim 16, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the set of processing core resources is further operable to determine the function of the machine learning model by: selecting one or more equations from a plurality of equations. See Col. 11, lines 50-55, (A machine learning model is considered to be one or more equations trained using a set of data. Users may interact with model training system 620 to provide data that causes the system to train one or more machine learning models, i.e. wherein the set of processing core resources is further operable to determine the function of the machine learning model by: selecting one or more equations from a plurality of equations (e.g. selecting a machine learning model for training comprises selecting a set of one or more equations represented as an ML Model).)
Regarding dependent claim 29,
As discussed above with claim 28, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the set of processing core resources is further operable to: identify the one or more equations within the model training operation. See Col. 12, lines 13-26, (Model training system 620 may package an algorithm that defines an ML model into a container image as part of a training process for the ML model. The algorithm may be pre-generated and obtained by a user from an algorithm repository, i.e. wherein the set of processing core resources is further operable to: identify the one or more equations within the model training operation. (e.g. the algorithm represents a process that is applied when the ML model is executed. An ML model may comprise one or more equations, therefore determining a model to execute and retrieving the associated algorithm requires retrieval and application of the one or more equations).)
Regarding dependent claim 30,
As discussed above with claim 29, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the set of processing core resources is further operable to determine the coefficients based on: identifying that the one or more equations includes the function. See Col. 5, lines 21-30, (ML model metadata may specify attributes indicating how an ML model is to be implemented, inputs, outputs, data types, basic instructions used by the model, etc.) See Col. 15, lines 7-15, (Model data may be generated which includes characteristics of the ML model being trained including coefficients of the ML model. Model data may also include information identifying the structure of an algorithm that defines an individual ML model, i.e. wherein the set of processing core resources is further operable to determine the coefficients based on: identifying that the one or more equations includes the function (e.g. a request to execute an ML model identifies a particular model to be executed wherein an ML model is associated with one or more equations and comprises metadata indicating instructions used by the model during execution. Executing a model includes determining coefficients).)
Regarding dependent claim 31,
As discussed above with claim 16, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the set of processing core resources is further operable to: identify the coefficients within the model training operation. See Col. 15, lines 8-15 & 35-38, (ML training container 630 may generate model data including coefficients associated with an ML model and stores the model data in a file system of the ML training container 630. A model data file may include model data information including a list of coefficients, i.e. wherein the set of processing core resources is further operable to: identify the coefficients within the model training operation.)
Regarding dependent claim 32,
As discussed above with claim 16, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the machine learning data corresponding to the dataset is data within the dataset or associated with the dataset. See Col. 15, lines 50-55. (Users may interact with model training system 620 to provide data that causes the system to train one or more ML models using a set of data.) See Col. 14, lines 6-12, (A virtual machine instance 622 may identify a type of training data indicated by a training request and select an ML model to train that corresponds with the identified type of training data, i.e. wherein the machine learning data corresponding to the dataset is data within the dataset or associated with the dataset.)
Regarding dependent claim 36,
As discussed above with claim 16, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong further discloses the step wherein the set of processing core resources is further operable to: store the function and the desired set of values for the coefficients as the machine learning model. See Col. 5, lines 7-14, (Storage service 108 may store a trained ML model that may be retrieved by a management module 116. Note Col. 15, lines 10-15 wherein model data includes characteristics of the machine learning model being trained, such as a number of layers in the machine learning model, hyperparameters of the machine learning model, coefficients of the machine learning model, weights of the machine learning model, etc., i.e. store the function and the desired set of values for the coefficients as the machine learning model.)
Regarding dependent claim 37,
The claim is analogous to the subject matter of dependent claim 28 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 38,
The claim is analogous to the subject matter of dependent claim 29 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 39,
The claim is analogous to the subject matter of dependent claim 30 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 40,
The claim is analogous to the subject matter of dependent claim 31 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 41,
The claim is analogous to the subject matter of dependent claim 32 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Claim(s) 33-34 and 42-43 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tong in view of MANIVANNAN and Ghassal as applied to claim 16 above, and further in view of Patel et al. (US PGPUB No. 2022/0382935; Pub. Date: Dec. 1, 2022).
Regarding dependent claim 33,
As discussed above with claim 16, Tong-MANIVANNAN-Ghassal discloses all of the limitations.
Tong-MANIVANNAN-Ghassal does not disclose the step wherein the set of processing core resources is further operable to identify the second set of processing core resources by: analyzing the function in light of the plurality of machine learning data to determine a level of processing.
Patel discloses the step wherein the set of processing core resources is further operable to identify the second set of processing core resources by: analyzing the function in light of the plurality of machine learning data to determine a level of processing. See Paragraph [0824], (Disclosing a system for receiving a simulated performance and presentation for a trial design of a set of trial designs. The system comprises a prediction engine 14701 configured to determine design progress data 14702 which may be received by a resource allocation engine, i.e. analyzing the function in light of the plurality of machine learning data to determine a level of processing (e.g. machine learning models may be trained on historical data to predict when resources will be needed.).)
Tong, MANIVANNAN, Ghassal and Patel are analogous art because they are in the same field of endeavor, machine learning techniques. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of Tong-MANIVANNAN-Ghassal to include the method of allocating processing resources based on a progress assessment as disclosed by Patel. Paragraph [0827] of Patel discloses that the system employs machine learning techniques capable of predicting when a computing resource will be available for use as well as monitoring for a determined trigger which would cause the system to allocate resources as soon as they are predicted to be needed. This represents an improvement in resource allocation.
Regarding dependent claim 34,
As discussed above with claim 33, Tong-MANIVANNAN-Ghassal-Patel discloses all of the limitations.
Patel further discloses the step wherein the set of processing core resources is further operable to:based on the level of processing and availability of the pluralities of processing core resources, identify the second plurality of processing core resources. See Paragraph [0824], (Polling engine 14712 is configured to identify a list of available resources including an availability, cost, computation capability, time to availability, etc. The resource allocation engine may receive data related to design progress 14702 within the platform indicating a current status of a workflow and may determine when additional resources should be requested such that they are available when needed, i.e. wherein the set of processing core resources is further operable to: based on the level of processing and availability of the pluralities of processing core resources, identify the second plurality of processing core resources.)
Regarding dependent claim 42,
The claim is analogous to the subject matter of dependent claim 33 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 43,
The claim is analogous to the subject matter of dependent claim 34 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Regarding dependent claim 45,
The claim is analogous to the subject matter of dependent claim 36 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Claim(s) 35 and 44 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tong in view of MANIVANNAN, Ghassal and Patel as applied to claim 16 above, and further in view of Stanley et al. (International Publication Number: WO 2023/048770 A1; International Publication Date: March 30, 2023).
Regarding dependent claim 35,
As discussed above with claim 34, Tong-MANIVANNAN-Ghassal-Patel discloses all of the limitations.
Tong-MANIVANNAN-Ghassal-Patel does not disclose the step wherein the set of processing core resources is further operable to: based on the level of processing, determine an overwrite level regarding overlapping processing of the plurality of machine learning training data.
Stanley discloses the step wherein the set of processing core resources is further operable to: based on the level of processing, determine an overwrite level regarding overlapping processing of the plurality of machine learning training data. See Paragraph [00172], (Disclosing a system for data collection balancing for sustainable storage. The system may execute an ML model 1068 to determine a redundancy of data based on uniqueness. The system may determine if there are duplicates or overlaps of values of data across datasets, i.e. wherein the set of processing core resources is further operable to: based on the level of processing, determine an overwrite level regarding overlapping processing of the plurality of machine learning training data (e.g. by determining a redundancy of data across datasets).)
Tong, MANIVANNAN, Ghassal, Patel and Stanley are analogous art because they are in the same field of endeavor, machine learning techniques. It would have been obvious to anyone having ordinary skill in the art before the effective filing date to modify the system of Tong-MANIVANNAN-Ghassal-Patel to include the method of assessing data overlap during machine learning processes as disclosed by Stanley. Paragraph [0077] of Stanley discloses that the system may enable improved performance, management, security and coordination function between entities and enable infrastructure offload and/or communications coordination functions.
Regarding dependent claim 44,
The claim is analogous to the subject matter of dependent claim 35 directed to a non-transitory, computer readable medium and is rejected under similar rationale.
Response to Arguments
Applicant’s arguments with respect to the previously presented set of claims have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Applicant’s amendments have modified the scope of the claimed invention, therefore any arguments relating to prior combinations of references are moot.
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).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Fernando M Mari whose telephone number is (571)272-2498. The examiner can normally be reached Monday-Friday 7am-4pm.
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, Ann J. Lo can be reached at (571) 272-9767. 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.
/FMMV/Examiner, Art Unit 2159
/ANN J LO/Supervisory Patent Examiner, Art Unit 2159
Prosecution Insights