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 . This action is made final.
This office action is in response to the amendments filed on February 25, 2026.
Claims 1, 12, and 15 have been amended.
Response to Amendment
The amendments filed February 25, 2026 has been entered. Claims 1-24 remain pending in the application.
Response to Arguments
Response to Obviousness/103
Applicant's arguments filed February 25, 2026 have been fully considered but they are not persuasive. Specifically:
Applicant Argues
Claims 1-10, 13-20, 23, and 24 stand rejected under 35 U.S.C. § 103 as being unpatentable over U.S. Pub. No. 2024/0044801 ("Zhang") in view of U.S. Pat. No. 11,430,548 ("Shannon"). Claims 11, 12, 21, and 22 stand rejected under 35 U.S.C. § 103 as being unpatentable over Zhang in view of Shannon and "LightGBM: A Highly Efficient Gradient Boosting Decision Tree" ("Ke"). These rejections are respectfully traversed.
Support includes at least [0140]-[0145] and FIG. 8 of the present application. The Office Action acknowledges that Zhang does not disclose a training dataset including first and second bins and instead relies on Shannon. However, Shannon does not teach the additional features of amended claim 1. Shannon merely describes using historical data 201 associated with a selected patient population to train a model. See Shannon, 8:3-27. Shannon is silent on selecting elements of the historical data, determining whether each element falls within or outside a specified range, assigning elements to first and second bins based on that determination, and applying under-sampling to the first bin and over-sampling to the second bin to generate updated bins, as required by amended claim 1.
For at least these reasons, the Applicant submits that claim 1 defines over the cited art.
Remaining Claims
Independent claim 15 includes similar features, and is patentable for at least similar reasons, as claim 1. The remaining claims are patentable at least by virtue of their dependency upon a patentable independent claim.
Examiner Response
Applicants’ arguments regarding the prior rejection have been considered to the extent the applicant argues that the previously applied references do not teach the newly added limitations, the arguments are moved in view of the new ground of rejection set forth below, which applies additional art to the amendment limitations period to the extent applicants’ arguments remain relevant to the references still applied, they are not perceived for the reasons discussed below.
Applicant argues that the previously applied references do not teach the newly added limitations directed to selecting elements of the training data set, determining whether each element is with an arranged, and assigning the elements to first and 2nd bids based on that determination. These arguments are moved and viewed the new ground rejection set forth herein, which applies additional art to the newly added Limitations. The new ground is an necessitated by applicants’ amendment.
Further come applicant’s arguments are not persuasive because they attack the references individually, while the rejection is based on the combined teachings of the applied references. Zhang is relied upon for predictive healthcare/pharmacy analytics using historical demographic and drug history data, selected input features, model training, and output predictions relating to drug events, enrollment/disenrollment, and plan/member outcomes. Shannon is relied upon for automated machine learning training, preprocessing, model quality/performance Evaluation, and hyperparameter optimization. Applicant has not shown that the combined teachings of the references fail to teachers suggest the claimed data preparation, model training, performance evaluation, and hyperparameter optimization workflow.
In particular Shannon is not relied upon for the healthcare/pharmacy prediction context top by Zhang and Zhang is not relied upon for the automated hyperparameter optimization teaching the Shannon. Rather, it would have been obvious to a combined Zang’s predictive health care modeling system was Shannon’s automated machine learning training and hyperparameter optimization techniques in order to improve model training, model selection, and predictive performance using known automated machine learning techniques.
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Response to Statutory Subject Matter /101
Applicant's arguments filed February 35, 026 have been fully considered but they are not persuasive. Specifically:
Applicant Argues
The Applicant respectfully submits that amended claim 1 does not recite an abstract idea. The Office Action alleges that claim 1, prior to amendment, falls within the "mental process" grouping of abstract ideas. However, claim 1, at least as amended, recites specific technical steps for training and optimizing a machine learning model that cannot be practically performed in the human mind.
The Applicant respectfully submits that amended claim 1 integrates any alleged abstract idea into a practical application. The claim recites a structured and computer-implemented process for training machine learning models, including generating an updated training dataset through targeted under-sampling and over- sampling and optimizing hyperparameters based on performance metrics. These concrete steps improve the quality of the training data and enhance model performance, thereby improving the operation of the machine learning system itself. See the published application at [0003], [0004], [0120], and [0145].
As recognized in the recent precedential decision Ex parte Desjardins1, improvements that enhance how a machine-learning model operate - such as reducing system complexity, improving efficiency, and maintaining model performance across tasks - constitute meaningful technical advancements. Further, the Director of the USPTO includes the following guidance to the PTAB and examining corps:
Categorically excluding Al innovations from patent protection in the
United States jeopardizes America's leadership in this critical emerging technology. Yet, under the panel's reasoning, many Al innovations are
potentially unpatentable-even if they are adequately described and
nonobvious-because the panel essentially equated any machine learning
with an unpatentable "algorithm" and the remaining additional elements as
"generic computer components," without adequate explanation. Dec. 24.
Examiners and panels should not evaluate claims at such a high level of
generality.
Ex Parte Desjardins, 9. Consistent with that reasoning, the claimed invention similarly improves model training and optimization, resulting in improved system operation rather than merely implementing an abstract idea on a generic computer. This optimization further demonstrates that claim 1 is directed to a practical application, not an abstract idea. For at least these reasons, the Applicant respectfully submits that claim 1 is directed to eligible subject matter.
Remaining Claims
Independent claim 15 includes similar features, and is directed to eligible subject matter for at least similar reasons, as claim 1. The remaining claims ultimately depend from claims 1 or 15 and are directed to eligible subject matter for at least the same reasons as the respective independent claims.
Examiner Response
Claim 1 continues to recite an abstract idea including collecting, evaluating, classifying, and manipulating data for machine learning model training and optimization. For example, claim one recite selecting elements of a training data set, determining whether each element is within a range, and assigning the elements to first and second bins based on whether each element is within or outside the range. These limitations amount to evaluating and classifying information based on a rural or condition.
The remaining limitations, including applying under sampling and oversampling techniques training a machine learning model, determining performance metrics, comparing the metrics to a threshold, and saving hyperparameters, are recited at a high level of generality. The claim does not recite a particular sampling algorithm, modeling architecture, training rule, Objective function, hardware arrangement, or other specific technological mechanism that improves the operation of a computer or machine learning model itself. Rather, The clean broadly recites using generic machine learning and data processing operations to prepare data, train a model, evaluate model performance, and select hyperparameters.
Applicant’s arguments that the claim improves training data quality and model performance Is not also not persuasive. The alleged improvement is stated functionally and generically, without acclaimed technical mechanism that improves computer functionality or the operation of the model itself. Unlike the claims at issue in Ex Parte Desjardins, amended claim 1 does not recite a specific machine learning improvement such as preserving prior task performance, reducing model storage, reducing system complexity, changing model architecture, or otherwise improving computer functionality. Instead, claim one recites generic range based binning, sampling, model training, performance evaluation, and hyperparameter selection.
Accordingly, amended claim 1 remains directed to an abstract idea and does not integrate the abstract idea into a practical application. The additional elements, considered individually and as an ordered combination, do not amount to significantly more than the abstract idea because they merely implement the abstract data processing and model training operations using generic computer components and a generic machine learning model. Therefore, the rejection of claim 1 under 35 USC Subsection 101 is maintained.
Independent claim 15 recites substantially similar limitations in system form and is rejected for at least the same reasons. The dependent claims do not cure the deficiencies of the independent claims. Accordingly, the rejection of claims 1-24 is maintained.
Response to Double Patenting
Applicant's arguments filed February 35, 026 have been fully considered but they are not persuasive. Specifically:
Applicant Argues
Claims 1-24 stand rejected on the ground of nonstatutory obviousness-type double patenting over U.S. Pub. No. 2024/0256987, over U.S. Pub. No. 2024/0256945 in view of "Care Episode Retrieval" ("Moen"), and over U.S. Pub. No. 2024/0256985 in view of Moen.
While the Applicant does not necessarily agree with the double patenting rejections, the Applicant has amended independent claims 1 and 15. Therefore, the Applicant respectfully requests reconsideration and withdrawal of the double patenting rejections.
Examiners Response
Applicant’s arguments regarding the double patenting rejections have been considered but are not persuasive. Applicant merely states that claims 1 and 15 have been amendment and requests reconsideration, but applicant has not filled a terminal disclaimer and has not identified any specific amended limitation that renders the pending claims patently distinct from the claims relied upon in the double patenting rejections. Accordingly, the double patent rejections are maintained.
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 therefor, subject to the conditions and requirements of this title.
To determine if a claim is directed to patent ineligible subject matter, the Court has guided the Office to apply the Alice/Mayo test, which requires:
Step 1: Determining if the claim falls within a statutory category.
Step 2A: Determining if the claim is directed to a patent ineligible judicial exception consisting of a law of nature, a natural phenomenon, or abstract idea; and Step 2A is a two prong inquiry. MPEP 2106.04(II)(A). Under the first prong, examiners evaluate whether a law of nature, natural phenomenon, or abstract idea is set forth or described in the claim. Abstract ideas include mathematical concepts, certain methods of organizing human activity, and mental processes. MPEP 2104.04(a)(2). The second prong is an inquiry into whether the claim integrates a judicial exception into a practical application. MPEP 2106.04(d).
Step 2B: If the claim is directed to a judicial exception, determining if the claim recites limitations or elements that amount to significantly more than the judicial exception. (See MPEP 2106).
Claims 1-24 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1-14 are directed to a method (a process), and Claim 15-24 are directed to a system (a machine). Therefore, Claims 1-24 are directed to a process, machine or manufacture or composition of matter.
Regarding claim 1
Step 2A Prong 1
Claim 1 recites the following mental processes, that in each case under the broadest reasonable interpretation, covers performance of the limitation in the mind (including observation, evaluation, judgement, opinion) or with the aid of pencil and paper but for recitation of generic computer components (e.g., “machine learning model”) [see MPEP 2106.04(a)(2)(III)].
“selecting a set of elements of the training data set” (e.g., selecting data from a table/list… etc.)
“determining whether the respective element is within a range” (e.g., observing selected data and comparing it to a predetermined parameter based on its value or content.)
“in response to the respective element being within the range, adding the respective element to the first bin” (e.g., identifying the selected data matched the predetermined variable and placing/setting it into a new list/table etc.)
“in response to the respective element being outside of the range, adding the respective element to the second bin” (e.g., identifying the selected data does not match the predetermined variable and placing/setting it into a new list/table etc.)
“generating an updated training data set by merging the updated first bin and the updated second bin” (e.g., combining two data sets a human can update a dataset from desired data form separate datasets)
“determining whether the baseline performance metrics are above a threshold” (e.g., a human can determine if a result is passing certain criteria)
Claim 1 recites the following mathematical process, that in each case under the broadest reasonable interpretation, involves mathematical relationships, formulas, calculations, or algorithms implemented using generic computer components (e.g., “machine learning model”) [see MPEP 2106.04(a)(2)(I)].
“applying an under-sampling technique to the elements of the first bin to generate an updated first bin” (e.g., a mathematical data transformation, interpolation of decreasing a data sample set. This can also be interpreted as a mental concept as it is merely reducing the size of a sample set randomly, which a person can do with aid of pen and paper.)
“applying an over-sampling technique to the elements of the second bin to generate an updated second bin” (e.g., a mathematical data transformation, interpolation of increasing a data sample set. This can also be interpreted as a mental concept as it is merely increasing the size of a sample set randomly, which a person can do with aid of pen and paper.)
Accordingly, at Step 2A, prong one, the claim is directed to an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “machine learning model” which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)). In particular, the recited “machine learning model” is merely a generic computer component, because it is recited to perform the function of implementing the “training data set” and “hyperparameters”, the claims do not recite any particular structure for how such “machine learning model” is implemented.
Regarding the “loading a training data set, wherein the training data set includes a first bin and a second bin”, “loading baseline hyperparameters”, “providing the updated training data set as inputs to the trained machine learning model configured with the baseline hyperparameters to determine baseline performance metrics”, and “providing input variables to the trained machine learning model configured with the optimal hyperparameters to generate output variables” limitations, these additional elements are recited at a high level of generality and amounts to extra-solution activity of transforming data to implement by loading the datasets in a machine learning model architecture, i.e. pre-solution activity of gathering data for use in the claimed process (i.e., loading the datasets to be manipulated in the architecture) and post-solution activity of gathering data(i.e., providing the manipulated datasets for further processing) (see MPEP 2106.05(g)).
Regarding the “training a machine learning model with the updated training data set”, “configuring the trained machine learning model with the baseline hyperparameters”, and “in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters” limitations, these additional elements are recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using a generic computer component (i.e., saving a template and saving a preferred template) (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional elements of a “machine learning model” which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using a generic computer component (See MPEP 2106.05(f)).
Regarding the loading a training data set, wherein the training data set includes a first bin and a second bin”, “loading baseline hyperparameters”, “providing the updated training data set as inputs to the trained machine learning model configured with the baseline hyperparameters to determine baseline performance metrics”, and “providing input variables to the trained machine learning model configured with the optimal hyperparameters to generate output variables” limitations, as discussed above, these additional elements are recited at a high level of generality and amounts to extra-solution activity of receiving/providing data, i.e. pre and post-solution activity of gathering data for use in the claimed process. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood, routine, and conventional (See MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory").
Regarding the “training a machine learning model with the updated training data set”, “configuring the trained machine learning model with the baseline hyperparameters”, and “in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters” limitations, these additional elements are recited at a high-level of generality and amounts to no more than adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. Accordingly, this additional element does not add significantly more than the judicial exception. (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Regarding claim 2
Step 2A Prong 1
Claim 2 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “the input variables include an identifier of an entity in a population”, “the output variables include a score for the entity indicated by the identifier “, and “the score indicates a likelihood of a feature of merit exceeding a threshold”, limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “the input variables include an identifier of an entity in a population”, “the output variables include a score for the entity indicated by the identifier “, and “the score indicates a likelihood of a feature of merit exceeding a threshold”, limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 3
Step 2A Prong 1
Claim 3 recites the following mathematical process, that in each case under the broadest reasonable interpretation, involves mathematical relationships, formulas, calculations, or algorithms implemented using generic computer components (e.g., “machine learning model”) [see MPEP 2106.04(a)(2)(I)].
“wherein the score is a value between zero and one hundred inclusive” (numeric normalization, representation or a number on a defined scale)
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
In accordance with Step 2A, Prong 2, the claim does not include any additional elements and the judicial exception is not integrated into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Regarding claim 4
Step 2A Prong 1
Claim 4 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “the population includes entities that consume services; and the feature of merit is a measure of service consumption”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “the population includes entities that consume services; and the feature of merit is a measure of service consumption”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 5
Step 2A Prong 1
Claim 5 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “the population includes entities that coordinate services”, and “the feature of merit is an amount of services”, limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “the population includes entities that coordinate services”, and “the feature of merit is an amount of services”, limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 6
Step 2A Prong 1
Claim 6 recites the following mental processes, that in each case under the broadest reasonable interpretation, covers performance of the limitation in the mind (including observation, evaluation, judgement, opinion) or with the aid of pencil and paper but for recitation of generic computer components (e.g., “model learning model”) [see MPEP 2106.04(a)(2)(III)].
“in response to determining that the baseline metrics are not above the threshold, adjusting the baseline hyperparameters” (e.g., a human can adjust the baseline parameters of a dataset if they are not meeting certain criteria)
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
In accordance with Step 2A, Prong 2, the claim does not include any additional elements and the judicial exception is not integrated into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Regarding claim 7
Step 2A Prong 1
Claim 7 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “configuring the machine learning model with the adjusted hyperparameters”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “configuring the machine learning model with the adjusted hyperparameters”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 8
Step 2A Prong 1
Claim 8 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “providing the training data set as inputs to the machine learning model configured with the adjusted hyperparameters to determine updated performance metrics”, limitation, the additional element is recited at a high-level of generality and amounts to extra-solution activity of obtaining data to input for a model, i.e., pre-solution activity of data gathering (e.g., obtaining information for processing in a computer system (see MPEP 2106.05(g)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “providing the training data set as inputs to the machine learning model configured with the adjusted hyperparameters to determine updated performance metrics”, limitation, the additional element is recited at a high-level of generality and amounts to extra-solution activity of obtaining data to input for a model, i.e., pre-solution activity of data gathering. The courts have found limitations directed to obtaining information electronically, recited at a high-level of generality, to be well-understood, routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory").
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 9
Step 2A Prong 1
Claim 9 recites the following mental processes, that in each case under the broadest reasonable interpretation, covers performance of the limitation in the mind (including observation, evaluation, judgement, opinion) or with the aid of pencil and paper but for recitation of generic computer components (e.g., “model learning model”) [see MPEP 2106.04(a)(2)(III)].
“determining whether the updated performance metrics are more optimal than the baseline performance metrics” (e.g., a human can compare performance of a metrics and determine if they are preferred in comparison of the baseline templates)
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
In accordance with Step 2A, Prong 2, the claim does not include any additional elements and the judicial exception is not integrated into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Regarding claim 10
Step 2A Prong 1
Claim 10 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “saving the adjusted hyperparameters as the baseline hyperparameters”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “saving the adjusted hyperparameters as the baseline hyperparameters”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 11
Step 2A Prong 1
Claim 11 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “wherein the machine learning model is a light gradient-boosting machine (LightGBM) regressor model”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “wherein the machine learning model is a light gradient-boosting machine (LightGBM) regressor model”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 12
Step 2A Prong 1
Claim 12 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “the output variables include at least one of (i) total drug costs for a set of months, (ii) member months for health insurance organizations, and (iii) total medical costs for the member”, “the output variables are stored in one or more databases”, and “the one or more databases are configured to feed into visualization software” limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “the output variables include at least one of (i) total drug costs for a months, (ii) member months for health insurance organizations, and (iii) total medical costs for the member”, “the output variables are stored in one or more databases”, and “the one or more databases feed into visualization software” limitations, which are recited at a high-level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 13
Step 2A Prong 1
Claim 13 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “wherein the input variables are stored on one or more storage devices”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “wherein the input variables are stored on one or more storage devices”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claim 14
Step 2A Prong 1
Claim 14 does not introduce any new abstract ideas, but recites the abstract idea identified in claim 1.
Accordingly, at Step 2A, prong one, the claim recites an abstract idea.
Step 2A Prong 2
The judicial exception is not integrated into a practical application. In particular, the claim recites the additional element of a “wherein the machine learning model is configured to access the input variables via one or more networks”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2A, prong two, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed above, the additional element of a “wherein the machine learning model is configured to access the input variables via one or more networks”, limitation, which is recited at a high-level of generality such that it amounts to no more than mere instructions to apply the exception using generic computer components (See MPEP 2106.05(f)).
Accordingly, at Step 2B, the additional element individually or in combination does not amount to significantly more than the judicial exception.
Regarding claims 15-24
Claims 15-24 recites a system comprising hardware components. This system corresponds directly to the method steps of claims 1, and 6-14, respectively, with the addition on hardware and computer-readable instructions which are insufficient to render the claims subject matter eligible for the same reasons described above.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claim(s) 1-10, 13-20, and 23-24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al. (US 20240044801 A1, referred to as Zhang) in view of Shannon et al. (US 11430548 B2, referred to as Shannon), in view of Branco et al. (“SMOGN: a Pre-processing Approach for Imbalanced Regression”, referred to as Branco).
Regarding claim 1, Zhang teaches, a computer-implemented method comprising ([0034] Describes that the discourse is a system and method to execute the instructions within.):
generating an updated training data set by merging the updated first bin and the updated second bin([0117]: Describes that after augmentation and preprocessing, the dataset is used to train models, merging augmented data and nonaugmented data implying a unified training dataset.);
loading baseline hyperparameters([0072]: Describes initializing a model with a starting set of hyperparameters before optimization, representing baseline values to be evaluated, teaching lading baseline hyperparameters.);
configuring the trained machine learning model with the baseline hyperparameters;
providing the updated training data set as inputs to the trained machine learning model configured with the baseline hyperparameters to determine baseline performance metrics ([0072]: Describes using Bayesian Optimization to choose model hyperparameters, for each candidate set (includes initial/baseline settings) the system configures the model with those hyperparameters and evaluates performance via cross-validation using metrices such as misclassification rate cross-entropy, or normalized MSE i.e., it provides the (updated) training data to the model configured with the baseline/candidate hyperparameters to determine baseline performance metrics. ;[0118]: Details the operational flow, of split training data into training/validation, optimize hyperparameters and during that process train/configure models with the chosen (baseline/candidate) hyperparameters and feed the split training data into those models to compute a quality measure (validation score) again, providing the updated dataset to a model configured with baseline hyperparameters to determine baseline metrics.);
determining whether the baseline performance metrics are above a threshold([0072]: Describes minimizing a score of the hyperparameters, which correspond to checking the criterion against a threshold.);
in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters([0092-0093]: Describes that after the Bayesian Optimization finds the best hyperparameters, the found hyperparameters are saved and used for final training.);
configuring the trained machine learning model with optimal hyperparameters([0093], [0118], and [0120]: Describes training models which utilize the found hyperparameters.); and
providing input variables to the trained machine learning model configured with the optimal hyperparameters to generate output variables([0121]]: Describes “applying or inputting data representative of the spectrum data to the trained models … to generate the substance identification information” which applies optimized models to new input data to get output predictions.).
Although Zhang teaches generating an updated training data set by merging the updated first bin and the updated second bin, loading baseline hyperparameters; configuring a machine learning model with the baseline hyperparameters, providing the updated training data set as inputs to the machine learning model configured with the baseline hyperparameters to determine baseline performance metrics, determining whether the baseline performance metrics are above a threshold; in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters; configuring the machine learning model with optimal hyperparameters, and providing input variables to the machine learning model configured with the optimal hyperparameters to generate output variables. It does not teach loading a training data set, wherein the training data set includes a first bin and a second bin, applying an under-sampling technique to elements of the first bin to generate an updated first bin, and applying an over-sampling technique to elements of the second bin to generate an updated second bin.
Shannon teaches, loading a training data set, wherein the training data set includes a first bin and a second bin (Col 8, lines 3-27: Describes loading and dividing plan data into separate subsets of populations, one subset of existing (real) embers and another subset of prospective or synthetic members. These corresponds to as subsets of data being distinct categories of training data as separate bins);
applying an under-sampling technique to the elements of the first bin to generate an updated first bin(Col 10, lines 19-37: Describes a dis-enrollment is realized as members leaving in the simulations, this modified member population is the reduced (under-sampled) set of real members, corresponding to an updated first bin.);
applying an over-sampling technique to the elements of the second bin to generate an updated second bin(Col 16, lines 25-61: Describes adding synthetic/virtual records by copying or composting form real members. This corresponds to over-sampling of the second population.);
training a machine learning model … (Col. 8 lines 43-67 cont. Col 9, lines 1-17: Describes selecting a model type to train, optimizing hyperparameters for the selected models, using all of the training data in the found hyperparameters to train the selective models and validating the train model using test data.);
It would have been obvious to one of ordinary skill in the art at the time of the claimed invention to have combined sampling and machine learning pipeline of Zhang with Shannon’s data balancing. Doing so would enable the system to rebalance or augment training data to enhance the automatic hyperparameter optimization. This would enhance model generation and stability when the training data is skewed or sparse.
Although Zhang in view of Shannon teaches, loading a training data set, wherein the training data set includes a first bin and a second bin; applying an under-sampling technique to the elements of the first bin to generate an updated first bin; applying an over-sampling technique to the elements of the second bin to generate an updated second bin; and training a machine learning model. They do not teach, selecting a set of elements of the training data set; for each respective element of the set of elements: determining whether the respective element is within a range; in response to the respective element being within the range, adding the respective element to the first bin; in response to the respective element being outside of the range, adding the respective element to the second bin and training a machine learning model with the updated training data set.
Branco teaches selecting a set of elements of the training data set;
for each respective element of the set of elements:
determining whether the respective element is within a range;
in response to the respective element being within the range, adding the respective element to the first bin; and
in response to the respective element being outside of the range, adding the respective element to the second bin (Pages 37-38 Section 2, Pages 39-40 Section 4, and page 41, Algorithm 1: Describes a SMOGN preprocessing algorithm for imbalance regression using a data set D having examples {xi, yi}, where Y is a target variable. It orders the examples of this data set according to the target variable value in partitions the examples based on a relevance function and threshold. Algorithm 1 forms BinsN form examples where φ(yi) < tR and forms BinsR from examples where φ(yi) ≥ tR. Teaching determining whether each example falls within the relevance/ target value range and assigning this example to a first bin (BinsN) or a second bin (BinsR) Based on that determination. Thus, teaching applying random under sampling to BinsN and applying oversampling to BinsR to generate a modified data set.)
training a machine learning model (As described above with Shannon.) with the updated training data set (Branco Pages 37-38 Section 2, Pages 39-40 Section 4, and page 41, Algorithm 1: Describes that its SMOGN algorithm forms normal and rare bins, applies random under sampling to the normal bins, applies oversampling to the rare bins, and outputs newD, a new modified data set.)
It would have been obvious to one of ordinary skill in the art at the time of the claimed invention to have combined machine learning training workflow the Brancos threshold/range based binning and sampling. Doing so would change the original data distribution before learning allowing for the system to learn on rare and interesting cases.
Regrading claim 2, Zhang in view of Shannon, in view of Branco teaches, the method of claim 1.
Zhang further teaches, the input variables include an identifier of an entity in a population;
the output variables include a score for the entity indicated by the identifier; and
the score indicates a likelihood of a feature of merit exceeding a threshold ([0117-0121]: Describes receiving input data corresponding to samples (spectra) identified by source and class, and providing that data as input to trained machine learning models to generate prediction results or identification scores. The outputs represent probabilities or quantitative measures (analyte presence, classification accuracy) for each sample. These teaches input variables that identify a given entity/sample and output variables that represent a score (prediction or likelihood) indicating whether a measurable feature meets or exceeds a specified criterion (feature of merit to threshold)).
Regarding claim 3, Zhang in view of Shannon, in view of Branco teaches, the method of claim 2.
Zhang further teaches, wherein the score is a value between zero and one hundred inclusive ([0120-0121]:Describe outputting prediction or identification scores expressed as percentages or normalized probabilities of class membership or analyte concentration levels, typically scaled from 0 to 100 percent.).
Regarding claim 4, Zhang in view of Shannon, in view of Branco teaches, the method of claim 2.
Shannon further teaches, the population includes entities that consume services(,); and the feature of merit is a measure of service consumption (Col 2, lines 35-52and Col 10 lines 46-67 cont. Col 11, lines 1-5: Describes plan members who actively consume prescription and medical services e.g., taking prescribed drugs , enrolling and disenrolling from plans. The system predicts their usage and discontinuation, which constitutes a measure of service consumption.).
Regarding claim 5, Zhang in view of Shannon, in view of Branco teaches, the population includes entities that coordinate services; and
the feature of merit is an amount of services (Shannon Col. 2, lines 9-28, and Col. 4, lines 32-35: Describes that a PBM and plan sponsor are entities coordinating services, they organize, adjudicate, and reimburse pharmacies, effectively managing service delivery. The amount of services corresponds to the adjudicated or reimbursed transactions, which the system aggregates as plan-level totals i.e., total drug spends, plan margin, and benefit allocations per member).
Regarding claim 6, Zhang in view of Shannon, in view of Branco teaches, the method of claim 1.
Zhang further teaches, in response to determining that the baseline metrics are not above the threshold, adjusting the baseline hyperparameters ([0072]: Describes that when the validation score (misclassification percent, MSE) does not meet desired performance, Bayesian Optimization iteratively adjusts the hyperparameters combination to minimize the cost function.).
Regarding claim 7, Zhang in view of Shannon, in view of Branco teaches, the method of claim 6.
Zhang further teaches, configuring the machine learning model with the adjusted hyperparameters ([0072]: Describes that after each hyperparameter update , the system configures and trains the model with the new (adjusted) hyperparameters to evaluate its performance.).
Regarding claim 8, Zhang in view of Shannon, in view of Branco teaches, the method of claim 7.
Zhang further teaches, providing the training data set as inputs to the machine learning model configured with the adjusted hyperparameters to determine updated performance metrics [0118]: Describes re-training the model using the same training/validation splits after each new hyperparameter set is selected and computing validation scores for that configuration.).
Regarding claim 9, Zhang in view of Shannon, in view of Branco teaches, the method of claim 8.
Zhang further teaches, determining whether the updated performance metrics are more optimal than the baseline performance metrics ([0092-0093]: Describes that Bayesian Optimization compares the current (updated) score with previous iterations and retains the combination yielding a lower cost or higher accuracy.).
Regarding claim 10, Zhang in view of Shannon, in view of Branco teaches, the method of claim 9.
Zhang further teaches, in response to determining that the updated performance metrics are more optimal than the baseline performance metrics, saving the adjusted hyperparameters as the baseline hyperparameters ([0092-0093]: Describes that once Bayesian Optimization finds an improved combination, the best (found) hyperparameters are stored and used for subsequent model training.).
Regarding claim 13, Zhang in view of Shannon, in view of Branco teaches, the method of claim 1.
Zhang further teaches, wherein the input variables are stored on one or more storage devices ([0117-0118]: Describes receiving or importing datasets from instruments or memory, implying storage on local or remote storage devices prior to loading into the machine learning pipeline.).
Regarding claim 14, Zhang in view of Shannon, in view of Branco teaches, the method of claim 13.
Zhang further teaches, wherein the machine learning model is configured to access the input variables via one or more networks([0117]: Describes obtaining data “from external data sources or instruments”, which implies network access or data transmission paths between devices. Standard machine learning system practices includes network retrieval form storage or sensors.).
Regarding claims 15-20, 23, and 24. Which recites substantially the same limitations as claims 1-10, 13 and 14. Claims 15-20, 23 and 24 further recites a system comprising: memory hardware configured to store instructions; and processing hardware configured to execute the instructions (Zhang, [0130]: Describes that the use a processing device comprising computer hardware to execute the instructions of the method steps to accomplish the desired training.) which performs the method steps of claims 1-10, 13 and 14, respectively, and are therefore rejected on the same premise.
Claim(s) 11, 12, 21, and 22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al. (US 20240044801 A1, referred to as Zhang) in view of Shannon et al. (US 11430548 B2, referred to as Shannon), in vie of Branco et al. (“SMOGN: a Pre-processing Approach for Imbalanced Regression”, referred to as Branco), in view of Ke et al.(“LightGBM: A Highly Efficient Gradient Boosting Decision Tree”, referred to as Ke).
Regrading claim 11, Zhang in view of Shannon, in view of Branco teaches, the method of claim 10.
Although Zhang in view of Shannon, in view of Branco teaches the method of claim 10 it does not teach, wherein the machine learning model is a light gradient-boosting machine (LightGBM) regressor model.
Ke teaches, wherein the machine learning model is a light gradient-boosting machine (LightGBM) regressor model (Page 1, Abstract “We call our new GBDT implementation with GOSS and EFB LightGBM. Our experiments on multiple public datasets show that, LightGBM speeds up the training process of conventional GBDT by up to over 20 times while achieving almost the same accuracy” defining the use of LightGBM for machine learning model training and efficiency. ).
It would have been obvious to one of ordinary skill in the art at the time of the claimed invention to have combined Zhang’s machine learning pipeline with Ke’s light gradient-boosting model. Doing so would enable the system to provide faster training and improved accuracy on larger datasets, enhancing computational efficiency without changing overall pipeline behavior.
Regarding claim 12, Zhang in view of Shannon, in view of Ke teaches the method of claim 11.
Shannon further teaches, the output variables include at least one of (i) total drug costs for a set of months, (ii) member months for health insurance organizations, and (iii) total medical costs for the member (Col. 12, lines 30-44: Describes that the system computes total drug costs and aggregated plan-level cost metrics corresponding to total drug costs and total medical costs. ; Col. 16, lines 25-61: Describes month by month enrollment simulation, corresponding to the standard member months output for a health insurance plan.);
the output variables are stored in one or more databases(Col. 12, lines8-12: Describes that the output data (synthetic or forecasted prescription drug events and costs) are stored.); and
the one or more databases feed into visualization software (Col. 13, lines 4-15 and Col 24, lines 65-67 cont. Col 25, lines 1-:17 Describes that the data (drug costs, plan-level margins, etc.) are fed into visualization software.).
It would have been obvious to one of ordinary skill in the art at the time of the claimed invention to have combined Zhang’s machine learning pipeline with Ke’s light gradient-boosting model and Shannon’s health-care cost prediction and visualization platform. Doing so would enable the system to apply its trained model to healthcare data to output and store cost metrics and visualize them for plan management, improving usability without affecting functionality.
Regarding claims 21, and 22. Which recites substantially the same limitations as claims 11, and 12. Claims 21 and 22 further recites a system comprising: memory hardware configured to store instructions; and processing hardware configured to execute the instructions (Zhang, [0130]: Describes that the use a processing device comprising computer hardware to execute the instructions of the method steps to accomplish the desired training.) which performs the method steps of claims 11, and 12, respectively, and are therefore rejected on the same premise.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 1-24 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-17, 19-21 and 25-27 of U.S. Patent No. US 20240256987 A1. Although the claims at issue are not identical, they are not patentably distinct from each other because they describe the same training methods for a model using under and over sampling of data bins, with hyperparameter optimization based on performance metrics.
This is a provisional nonstatutory double patenting rejection because the patentably indistinct claims have not in fact been patented.
Instant Application
US 20240256987 A1
Claim 1
A computer-implemented method comprising: loading a training data set, wherein the training data set includes a first bin and a second bin; selecting a set of elements of the training data set; for each respective element of the set of elements: determining whether the respective element is within a range; in response to the respective element being within the range, adding the respective element to the first bin; and in response to the respective element being outside of the range, adding the respective element to the second bin; applying an under-sampling technique to the elements of the first bin to generate an updated first bin; applying an over-sampling technique to the elements of the second bin to generate an updated second bin; generating an updated training data set by merging the updated first bin and the updated second bin; training a machine learning model with the updated training data set; loading baseline hyperparameters; configuring a machine the trained machine learning model with the baseline hyperparameters; providing the updated training data set as inputs to the trained machine learning model configured with the baseline hyperparameters to determine baseline performance metrics; determining whether the baseline performance metrics are above a threshold; in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters; configuring the trained machine learning model with optimal hyperparameters; and providing input variables to the trained machine learning model configured with the optimal hyperparameters to generate output variables.
Claim 1
A non-transitory computer-readable medium comprising executable instructions for training and optimizing machine learning models, wherein the executable instructions include:
loading a training data set, wherein the training data set includes a first bin and a second bin;
applying an under-sampling technique to elements of the first bin to generate an updated first bin;
applying an over-sampling technique to elements of the second bin to generate an updated second bin;
generating an updated training data set by merging the updated first bin and the updated second bin;
loading baseline hyperparameters;
configuring a machine learning model with the baseline hyperparameters;
providing the updated training data set as inputs to the machine learning model configured with the baseline hyperparameters to determine baseline performance metrics;
determining whether the baseline performance metrics are above a threshold;
in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters;
configuring the machine learning model with optimal hyperparameters; and
providing input variables to the machine learning model configured with the optimal hyperparameters to generate output variables.
Claim 2
Claim 2
Claim 3
Claim 3
Claim 4
Claim 4
Claim 5
Claim 5
Claim 6
Claim 6
Claim 7
Claim 7
Claim 8
Claim 8
Claim 9
Claim 9
Claim 10
Claim 10
Claim 11
Claim 11
Claim 12
Claim 12
Claim 13
Claim 13
A system comprising:
memory hardware configured to store instructions; and
processing hardware configured to execute the instructions, wherein the instructions include:
loading input variables,
loading a first trained machine learning model,
providing input variables to the first trained machine learning model to generate first output variables,
determining whether the first output variables are above a threshold,
in response to determining that the first output variables are above the threshold:
loading a second trained machine learning model, and
providing the input variables to the second trained machine learning model to generate second output variables, and
in response to determining that the first output variables are not above the threshold:
loading a third trained machine learning model, and
providing the input variables to the third trained machine learning model to generate third output variables.
Claim 14
Claim 14
A system comprising: memory hardware configured to store instructions; and processing hardware configured to execute the instructions, wherein the instructions include: loading a training data set, wherein the training data set includes a first bin and a second bin; selecting a set of elements of the training data set; for each respective element of the set of elements: determining whether the respective element is within a range; in response to the respective element being within the range, adding the respective element to the first bin; and in response to the respective element being outside of the range, adding the respective element to the second bin; applying an under-sampling technique to the elements of the first bin to generate an updated first bin; applying an over-sampling technique to the elements of the second bin to generate an updated second bin; generating an updated training data set by merging the updated first bin and the updated second bin; training a machine learning model with the updated training data set; loading baseline hyperparameters; configuring a machine the trained machine learning model with the baseline hyperparameters; providing the updated training data set as inputs to the trained machine learning model configured with the baseline hyperparameters to determine baseline performance metrics; determining whether the baseline performance metrics are above a threshold; in response to determining that the baseline performance metrics are above the threshold, saving the baseline hyperparameters as optimal hyperparameters; configuring the trained machine learning model with optimal hyperparameters; and providing input variables to the trained machine learning model configured with the optimal hyperparameters to generate output variables.
Claim 15
Claim 16
Claim 16
Claim 17
Claim 17
Claim 18
Claim 19
Claim 19
Claim 20
Claim 20
Claim 21
Claim 21
Claim 25
Claim 22
Claim 26
Claim 23
Claim 27
Claim 24
Claims 1-24 provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-4, 7-8, and 11-27 of copending Application No. US 20240256945 A1 in view of Moen et al. ("Care Episode Retrieval", referred to as Moen). Both describe a system and method for loading over and under sampling, merging data, configuring model with baseline hyperparameters determine performance metrics relative to a threshold and adjusting or saving hyperparameters accordingly
This is a provisional nonstatutory double patenting rejection.
Instant Application
Patent No. US 20240256945 A1
The Instant application fails to particularly teach tokenization, frequency filtering and episode archiving. However, Moen teaches teach tokenization(pg. 120-121, Experiments), frequency filtering(pg. 120, Computing care episode similarity) and episode archiving(pg. 116, Introduction).
It would have been obvious to a person of ordinary skill in the arts at the times of the applicant’s invention to modify the teachings of the instant application by incorporating tokenization, frequency filtering and episode archiving as taught by Moen’s for the purpose of data preparation to train and optimize the model performance and ensure consistent feature encoding.
Claim 2
Claim 2
Claim 3
Claim 3
Claim 4
Claim 4
Claim 5
Claim 7
Claim 6
Claim 8
Claim 7
Claim 11
Claim 8
Claim 12
Claim 9
Claim 13
Claim 10
Claim 14
Claim 11
Claim 15
Claim 12
Claim 16
Claim 13
Claim 17
Claim 14
Claim 18
Claim 15
Claim 19
Claim 16
Claim 20
Claim 17
Claim 21
Claim 18
Claim 22
Claim 19
Claim 23
Claim 20
Claim 24
Claim 21
Claim 25
Claim 22
Claim 26
Claim 23
Claim 27
Claim 24
Claims 1-24 provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-2, 7-14, and 17-20 of copending Application No. US 20240256985 A1 in view of Moen et al. ("Care Episode Retrieval", referred to as Moen). Both describe a system and method for loading over and under sampling, merging data, configuring model with baseline hyperparameters determine performance metrics relative to a threshold and adjusting or saving hyperparameters accordingly
This is a provisional nonstatutory double patenting rejection.
Instant Application
Patent No. US 20240256985 A1
Claim 1
The Instant application fails to particularly teach tokenization, frequency filtering and episode archiving. However, Moen teaches teach tokenization(pg. 120-121, Experiments), frequency filtering(pg. 120, Computing care episode similarity) and episode archiving(pg. 116, Introduction).
It would have been obvious to a person of ordinary skill in the arts at the times of the applicant’s invention to modify the teachings of the instant application by incorporating tokenization, frequency filtering and episode archiving as taught by Moen’s for the purpose of data preparation to train and optimize the model performance and ensure consistent feature encoding.
Claim 2
Claim 2
Claim 3
Claim 7
Claim 4
Claim 8
Claim 5
Claim 9
Claim 6
Claim 10
Claim 7
Claim 11
Claim 8
Claim 12
Claim 9
Claim 13
Claim 10
Claim 14
Claim 11
Claim 17
Claim 12
Claim 18
Claim 13
Claim 19
Claim 14
Claim 20
Claim 15
Claim 16
Claim 17
Claim 18
Claim 19
Claim 20
Claim 21
Claim 22
Claim 23
Claim 24
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 DONALD T RODEN whose telephone number is (571)272-6441. The examiner can normally be reached Mon-Thur 8:00-5:00 EST.
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, Omar Fernandez Rivas can be reached at (571) 272-2589. 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.
/D.T.R./Examiner, Art Unit 2128
/OMAR F FERNANDEZ RIVAS/Supervisory Patent Examiner, Art Unit 2128
Prosecution Insights