CTFR 18/057,080 CTFR 100396 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia 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 in response to amendments filed February 9 th , 2026. The status of the claims is as follows. Claims 1-2, 4, 6-14, 17-20 are amended. Claims 3, 5, 15-16 are cancelled. Claims 1-2, 4, 6-14, 17-20 are currently pending. Claim Rejections - 35 USC § 103 07-20-aia AIA 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. 07-23-aia AIA The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. Claims 1-2, 6-14, 17-20 are rejected under 35 U.S.C. 103 as being anticipated by Shahhosseini et al. (“Improved weighted random forest for classification problems” [2020], hereinafter “Shahhosseini”) in view of Rendon et al. (“Structural Combination of Neural Network Models” [2016], hereinafter “Rendon”). Regarding Claim 1, Shahhosseini discloses A multi-label ensemble learning method performed by a multi-label ensemble learning apparatus including a memory storing a plurality of prediction models and a processor, the method comprising: collecting, by the processor, a prediction value for learning data for each of a plurality of prediction models … ; calculating, by the processor, a prediction error of each of the prediction models using the prediction value of each of the prediction models and a correct answer prediction value; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein the learning of the random forest ensemble method comprising multiple class labels reads on a multi-label ensemble learning method; Wherein the out-of-bag predictions Yj read on collected prediction values for each of the j plurality of prediction models; Wherein the calculated accuracy proportion of correct predictions of all j decision trees against the actual response values Y reads on inherently calculating a prediction error between Yj and Y of each of the prediction models (since accuracy is calculated upon all j trees)) generating, by the processor, a weight label for each of the prediction models based on the prediction error, wherein the weight label is generated by the processor, to control contribution or each of the prediction models to an ensemble prediction according to the prediction error; and learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (Shahhosseini [Figure 2]; PNG media_image2.png 423 511 media_image2.png Greyscale Wherein the generated weights w1, w2 … wn based on the optimization model process read on generated weight labels for each of the prediction models based on the prediction error; wherein the optimization model learning being used to optimize the random forest classifier (ensemble weight prediction model) reads on learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (P = round(sum(wiPi) essentially applying the weight label to generate new learned weights); wherein the weight labels generated by the processor dictating the contribution of each decision tree to the ensemble prediction thus reads on controlled contribution of each of the prediction models to an ensemble prediction) wherein the generating the weight label includes: calculating, by the processor, error-based weight scores for the prediction models based on the prediction error; and generating, by the processor, the weight label based on the error-based weight scores; (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein the optimization model step calculating PNG media_image4.png 45 258 media_image4.png Greyscale as an intermediary step reads on calculation of weight scores wjyj + 0.5 for the prediction models; wherein the max optimization function comparing the accuracies of the weight scores reads on iterations of weight scores and their associated accuracies being optimized when compared against other iterations of weight scores and their associated accuracies, thus reading on the weight scores being inherently error-based since they are being optimized in regards to the error and constraints; Wherein w1, w2 … wn generated from the optimization model step comprising error-based weight scores read on generated weight labels based on the error-based weight scores) optionally selecting, by the processor, at least some of the error-based weight scores for the prediction models; and generating, by the processor, weight labels for prediction models corresponding to the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein the selection of all j = 1 … k error-based weight score for the prediction models reads on optionally selecting at least some of error-based weight scores for the prediction models Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein w1, w2 … wn generated from the optimization model step comprising error-based weight scores read on generated weight labels based on the error-based weight scores) Shahhosseini fails to disclose but Rendon discloses wherein the plurality of prediction models generates future state time-series prediction values using the learning data including past state time- series data; (Rendon [Abstract]; “Forecasts combinations normally use point forecasts that were obtained from different models or sources ([1], [2], [3]). This paper explores the incorporation of internal structure parameters of feed-forward neural network (NN) models as an approach to combine their forecasts via ensembles. First, the generated NN models that could be part of the ensembles are subject to a clustering algorithm that uses the structure parameters and, from each of the clusters obtained, a small set of models is selected and their forecasts are combined in a two-stage procedure. Secondly, in an alternative and simpler implementation, a subset of the generated NN models is selected by using several reference points in the model structure parameter space. The choice of the reference points is optimised through a genetic algorithm and the models selected are averaged. Hourly electricity demand time series is used to assess multi-step ahead forecasting performance for up to a 12 hours horizon. Results are compared against several statistical benchmarks, the average of the individual forecasts and the best models in the ensembles.” wherein forecast predictions being measured and trained through hourly electricity demand time series thus reads on the learning data and generated prediction values including past state time-series data in inference) wherein, prediction models not corresponding to the at least some optionally selected error-based weight scores are excluded from the ensemble prediction (Rendon [Abstract]; “Forecasts combinations normally use point forecasts that were obtained from different models or sources ([1], [2], [3]). This paper explores the incorporation of internal structure parameters of feed-forward neural network (NN) models as an approach to combine their forecasts via ensembles. First, the generated NN models that could be part of the ensembles are subject to a clustering algorithm that uses the structure parameters and, from each of the clusters obtained, a small set of models is selected and their forecasts are combined in a two-stage procedure. Secondly, in an alternative and simpler implementation, a subset of the generated NN models is selected by using several reference points in the model structure parameter space. The choice of the reference points is optimised through a genetic algorithm and the models selected are averaged. Hourly electricity demand time series is used to assess multi-step ahead forecasting performance for up to a 12 hours horizon. Results are compared against several statistical benchmarks, the average of the individual forecasts and the best models in the ensembles” wherein the selected subset of generated NN models being used for averaging to determine an ensemble product implicitly reads on the exclusion of prediction models outside of that select subset). It would have been obvious to modify Shahhosseini’s ensemble model to incorporate Rendon’s method of completely excluding models not corresponding to select performances in a time-series environment. One would have been motivated to do so because “conducted model selection with strategies based on sequential statistical tests, information criteria and cross validation … identify an appropriate network configuration” thus allowing Shahhosseini’s ensemble model to only comprise optimal appropriate models to improve network configuration. Regarding Claim 2, Shahhosseini/Rendon teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Shahhosseini/Rendon further discloses wherein the learning the ensemble weight prediction model comprises learning, by the processor, the ensemble weight prediction model so that the weight of each of the prediction models and the weight label are minimized; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein the learned ensemble prediction model comprising weights of each of the prediction models and the weight label being minimized thus reads on all weights being set to the same minimized value. As such, the disclosure of the summation of all weights adding up to 1 inherently reads on each of the weights and weight labels being an equal minimized value) Regarding Claim 6, Shahhosseini/Rendon teaches the method of Claim 1 (and thus the rejection of Claim 5 is incorporated). Shahhosseini/Rendon further discloses generating the weight label of each of the prediction models comprises generating, by the processor, the weight label of each of the prediction models, by setting a sum of the at least some optionally selected error-based weight scores to 1 and setting, by the processor, the remaining error-based weight scores to 0 through a normalization process for the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein normalizing the second argument of the accuracy() function based variable to be rounded to the nearest class labels 0 and 1 reads on the optimization model step comprising a normalization process wherein some k error-based weight scores are set to 1 while the remaining are set to 0) Regarding Claim 7, Shahhosseini/Rendon teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Shahhosseini/Rendon further discloses optionally selecting at least some of the error-based weight scores comprises optionally selecting, by the processor, at least some of the error-based weight scores for the prediction models using a predetermined second parameter value; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Shahhosseini [Figure 2]; PNG media_image6.png 418 501 media_image6.png Greyscale Wherein the parameter value k reads on the selection of predetermined k error-based weight scores PNG media_image7.png 45 123 media_image7.png Greyscale following optimization constraints) Regarding Claim 8, Shahhosseini/Rendon teaches the method of Claim 7 (and thus the rejection of Claim 7 is incorporated). Shahhosseini/Rendon further discloses determining, by the processor, the number of prediction models using the second parameter value and the error-based weight scores for the prediction models and selecting, by the processor, an error-based weight score of a high value corresponding to the determined number of prediction models as the at least some error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Wherein the parameter value k inherently reads on the number of prediction models using the second parameter value and the error-based weight scores for the prediction models (j = 1 … k); wherein selection of error-based weight scores wjyj for all j = 1 … k determined number of prediction models that maximize the accuracy (prediction error) thus reads on selection of an error-based weight score of a high accuracy value) Regarding Claim 9, Shahhosseini/Rendon teaches the method of Claim 8 (and thus the rejection of Claim 8 is incorporated). Shahhosseini/Rendon further discloses calculating, by the processor, a normalization threshold using the second parameter value and the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale wherein PNG media_image7.png 45 123 media_image7.png Greyscale implicitly reads on the calculated normalization threshold of wjyj > 0 resulting in the class label 1, wjyj <=0 resulting in the class label 0. Since 0.5 results in a class label of 0, any of the k instances wherein wjyj is positive (wjyj > 0), the normalization will result in class label 1. Any of the k instances wherein wjyj is negative (wjyj < 0), the normalization will result in class label 0. As a result, the floor function applied to the summation series of the k error-based weight scores is thus interpreted as a calculated normalization threshold ) generating, by the processor, the weight label of each of the prediction models using the normalization threshold, the second parameter value and the error-based weight scores for the prediction models (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein the weight labels w1, w2 … wn generated through the optimization model step comprising PNG media_image7.png 45 123 media_image7.png Greyscale thus reads on generating weight labels for each of the prediction models through the normalization threshold, second parameter value, and error-based weight scores for the prediction models) Regarding Claim 10, Shahhosseini discloses collecting, by the processor, a prediction value for learning data of each of prediction models …; calculating, by the processor, a prediction error of each of the prediction models by comparing the prediction value of each of the prediction models and a correct answer prediction value; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein the learning of the random forest ensemble method comprising multiple class labels reads on a multi-label ensemble learning method; Wherein the out-of-bag predictions Yj read on collected prediction values for each of the j plurality of prediction models; Wherein the calculated accuracy proportion of correct predictions of all j decision trees against the actual response values Y reads on inherently calculating a prediction error between Yj and Y of each of the prediction models (since accuracy is calculated upon all j trees)) calculating, by the processor, error-based weight scores for the prediction models based on the prediction error; (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein the optimization model step calculating PNG media_image4.png 45 258 media_image4.png Greyscale as an intermediary step reads on calculation of weight scores wjyj + 0.5 for the prediction models; wherein the max optimization function comparing the accuracies of the weight scores reads on iterations of weight scores and their associated accuracies being optimized when compared against other iterations of weight scores and their associated accuracies, thus reading on the weight scores being inherently error-based since they are being optimized in regards to the error and constraints) wherein the error-based weight scores are calculated by the processor, to control contribution of each of the prediction models to an ensemble prediction according to the prediction error (Shahhosseini [Figure 2]; PNG media_image2.png 423 511 media_image2.png Greyscale Wherein the generated weights w1, w2 … wn based on the optimization model process read on generated weight labels for each of the prediction models based on the prediction error; wherein the optimization model learning being used to optimize the random forest classifier (ensemble weight prediction model) reads on learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (P = round(sum(wiPi) essentially applying the weight label to generate new learned weights); wherein the weight labels generated by the processor dictating the contribution of each decision tree to the ensemble prediction thus reads on controlled contribution of each of the prediction models to an ensemble prediction) optionally selecting at least some of the error-based weight scores for the prediction models using a predetermined parameter value; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Shahhosseini [Figure 2]; PNG media_image6.png 418 501 media_image6.png Greyscale Wherein the parameter value k reads on the selection of predetermined k error-based weight scores PNG media_image7.png 45 123 media_image7.png Greyscale following optimization constraints) and learning an ensemble weight prediction model for predicting a weight of each of the prediction models based on the at least some optionally selected error-based weight scores (Shahhosseini [Figure 2]; PNG media_image2.png 423 511 media_image2.png Greyscale wherein the optimization model learning being used to optimize the random forest classifier (ensemble weight prediction model) reads on learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (P = round(sum(wiPi) essentially applying the weight label to generate new learned weights)) Shahhosseini fails to disclose but Rendon discloses wherein the plurality of prediction models generates future state time-series prediction values using the learning data including past state time-series data; (Rendon [Abstract]; “Forecasts combinations normally use point forecasts that were obtained from different models or sources ([1], [2], [3]). This paper explores the incorporation of internal structure parameters of feed-forward neural network (NN) models as an approach to combine their forecasts via ensembles. First, the generated NN models that could be part of the ensembles are subject to a clustering algorithm that uses the structure parameters and, from each of the clusters obtained, a small set of models is selected and their forecasts are combined in a two-stage procedure. Secondly, in an alternative and simpler implementation, a subset of the generated NN models is selected by using several reference points in the model structure parameter space. The choice of the reference points is optimised through a genetic algorithm and the models selected are averaged. Hourly electricity demand time series is used to assess multi-step ahead forecasting performance for up to a 12 hours horizon. Results are compared against several statistical benchmarks, the average of the individual forecasts and the best models in the ensembles.” wherein forecast predictions being measured and trained through hourly electricity demand time series thus reads on the learning data and generated prediction values including past state time-series data in inference) wherein, prediction models not corresponding to the at least some optionally selected error-based weight scores are excluded from the ensemble prediction (Rendon [Abstract]; “Forecasts combinations normally use point forecasts that were obtained from different models or sources ([1], [2], [3]). This paper explores the incorporation of internal structure parameters of feed-forward neural network (NN) models as an approach to combine their forecasts via ensembles. First, the generated NN models that could be part of the ensembles are subject to a clustering algorithm that uses the structure parameters and, from each of the clusters obtained, a small set of models is selected and their forecasts are combined in a two-stage procedure. Secondly, in an alternative and simpler implementation, a subset of the generated NN models is selected by using several reference points in the model structure parameter space. The choice of the reference points is optimised through a genetic algorithm and the models selected are averaged. Hourly electricity demand time series is used to assess multi-step ahead forecasting performance for up to a 12 hours horizon. Results are compared against several statistical benchmarks, the average of the individual forecasts and the best models in the ensembles” wherein the selected subset of generated NN models being used for averaging to determine an ensemble product implicitly reads on the exclusion of prediction models outside of that select subset). It would have been obvious to modify Shahhosseini’s ensemble model to incorporate Rendon’s method of completely excluding models not corresponding to select performances in a time-series environment. One would have been motivated to do so because “conducted model selection with strategies based on sequential statistical tests, information criteria and cross validation … identify an appropriate network configuration” thus allowing Shahhosseini’s ensemble model to only comprise optimal appropriate models to improve network configuration. Regarding Claim 11, Shahhosseini/Rendon teaches the method of Claim 10 (and thus the rejection of Claim 10 is incorporated). Shahhosseini/Rendon further discloses generating, by the processor, the weight label of each of the prediction models, by setting a sum of the at least some optionally selected error-based weight scores to 1 and setting, by the processor, the remaining error-based weight scores to 0 through a normalization process for the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image1.png 342 491 media_image1.png Greyscale Wherein normalizing the second argument of the accuracy() function based variable to be rounded to the nearest class labels 0 and 1 reads on the optimization model step comprising a normalization process wherein some k error-based weight scores are set to 1 while the remaining are set to 0) wherein the learning the ensemble weight prediction model comprises learning, by the processor, the ensemble weight prediction model using the weight label of each of the prediction models (Shahhosseini [Figure 2]; PNG media_image2.png 423 511 media_image2.png Greyscale Wherein the generated weights w1, w2 … wn based on the optimization model process read on generated weight labels for each of the prediction models based on the prediction error; wherein the optimization model learning being used to optimize the random forest classifier (ensemble weight prediction model) reads on learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (P = round(sum(wiPi) essentially applying the weight label to generate new learned weights)) Regarding Claim 12, Shahhosseini/Rendon teaches the method of Claim 10 (and thus the rejection of Claim 10 is incorporated). Shahhosseini/Rendon further discloses determining, by the processor, the number of prediction models using the parameter value and the error-based weight scores for the prediction models and selecting, by the processor, an error-based weight score of a high value corresponding to the determined number of prediction models as the at least some error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Wherein the parameter value k inherently reads on the number of prediction models using the parameter value and the error-based weight scores for the prediction models (j = 1 … k); wherein selection of error-based weight scores wjyj for all j = 1 … k determined number of prediction models that maximize the accuracy (prediction error) thus reads on selection of an error-based weight score of a high accuracy value) Regarding Claim 13, Shahhosseini/Rendon teaches the method of Claim 12 (and thus the rejection of Claim 12 is incorporated). Shahhosseini/Rendon further discloses calculating, by the processor, a normalization threshold using the parameter value and the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale wherein PNG media_image7.png 45 123 media_image7.png Greyscale implicitly reads on the calculated normalization threshold of wjyj > 0 resulting in the class label 1, wjyj <=0 resulting in the class label 0. Since 0.5 results in a class label of 0, any of the parameter value k instances wherein wjyj is positive (wjyj > 0), the normalization will result in class label 1. Any of the k instances wherein wjyj is negative (wjyj < 0), the normalization will result in class label 0. As a result, the floor function applied to the summation series of the k error-based weight scores is thus interpreted as a calculated normalization threshold ) generating, by the processor, the weight label of each of the prediction models using the normalization threshold, the parameter value and the error-based weight scores for the prediction models (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein the weight labels w1, w2 … wn generated through the optimization model step comprising PNG media_image7.png 45 123 media_image7.png Greyscale thus reads on generating weight labels for each of the prediction models through the normalization threshold, parameter value, and error-based weight scores for the prediction models) wherein the learning the ensemble weight prediction mode comprises learning, by the processor, the ensemble weight prediction model using the weight label of each of the prediction models (Shahhosseini [Figure 2]; PNG media_image2.png 423 511 media_image2.png Greyscale Wherein the generated weights w1, w2 … wn based on the optimization model process read on generated weight labels for each of the prediction models based on the prediction error; wherein the optimization model learning being used to optimize the random forest classifier (ensemble weight prediction model) reads on learning an ensemble weight prediction model for predicting a weight of each of the prediction models using the weight label (P = round(sum(wiPi) essentially applying the weight label to generate new learned weights)) Claims 14 and 17 recite an apparatus comprising a collection unit, a generation unit, and a learning unit to execute the exact method of Claims 1and 6 respectively. Thus, Claims 14 and 17 is rejected for reasons set forth in the rejection of Claim 1 and 6 respectively. Regarding Claim 18, Shahhosseini/Rendon teaches the method of Claim 14 (and thus the rejection of Claim 14 is incorporated). Shahhosseini/Rendon further discloses wherein the processor is configured to optionally select at least some of the error-based weight scores for the prediction models using a predetermined parameter value; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Shahhosseini [Figure 2]; PNG media_image6.png 418 501 media_image6.png Greyscale Wherein the parameter value k reads on the selection of predetermined k error-based weight scores PNG media_image7.png 45 123 media_image7.png Greyscale following optimization constraints) Regarding Claim 19, Shahhosseini/Rendon teaches the method of Claim 18 (and thus the rejection of Claim 18 is incorporated). Shahhosseini/Rendon further discloses to determine the number of prediction models using the parameter value and the error-based weight scores for the prediction models and to select an error-based weight score of a high value corresponding to the determined number of prediction models as the at least some error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale Wherein the parameter value k inherently reads on the number of prediction models using the second parameter value and the error-based weight scores for the prediction models (j = 1 … k); wherein selection of error-based weight scores wjyj for all j = 1 … k determined number of prediction models that maximize the accuracy (prediction error) thus reads on selection of an error-based weight score of a high accuracy value) Regarding Claim 20, Shahhosseini/Rendon teaches the method of Claim 19 (and thus the rejection of Claim 19 is incorporated). Shahhosseini/Rendon further discloses to calculate a normalization threshold using the parameter value and the at least some optionally selected error-based weight scores; (Shahhosseini [Page 5]; PNG media_image5.png 417 518 media_image5.png Greyscale wherein PNG media_image7.png 45 123 media_image7.png Greyscale implicitly reads on the calculated normalization threshold of wjyj > 0 resulting in the class label 1, wjyj <=0 resulting in the class label 0. Since 0.5 results in a class label of 0, any of the parameter value k instances wherein wjyj is positive (wjyj > 0), the normalization will result in class label 1. Any of the k instances wherein wjyj is negative (wjyj < 0), the normalization will result in class label 0. As a result, the floor function applied to the summation series of the k error-based weight scores is thus interpreted as a calculated normalization threshold ) and to generate the weight label of each of the prediction models using the normalization threshold, the parameter value and the error-based weight scores for the prediction models; (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale Wherein the weight labels w1, w2 … wn generated through the optimization model step comprising PNG media_image7.png 45 123 media_image7.png Greyscale thus reads on generating weight labels for each of the prediction models through the normalization threshold, parameter value, and error-based weight scores for the prediction models) 07-21-aia AIA Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Shahhosseini et al. (“Improved weighted random forest for classification problems” [2020], hereinafter “Shahhosseini”) in view of Rendon et al. (“Structural Combination of Neural Network Models” [2016], hereinafter “Rendon”) in view of Elmachtoub et al. (“Decision trees for decision-making under the predict-then-optimize framework” [2020], hereinafter “Elmachtoub”) . Regarding Claim 4, Shahhosseini/Rendon teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). Shahhosseini/Rendon already discloses calculating, by the processor, the error based weight scores for the prediction models based on the prediction error (Shahhosseini [Figure 2]; PNG media_image3.png 122 483 media_image3.png Greyscale PNG media_image8.png 37 106 media_image8.png Greyscale Wherein the optimization model step calculating PNG media_image4.png 45 258 media_image4.png Greyscale as an intermediary step reads on calculation of weight scores wjyj + 0.5 for the prediction models; wherein the max optimization function comparing the accuracies of the weight scores reads on iterations of weight scores and their associated accuracies being optimized when compared against other iterations of weight scores and their associated accuracies, thus reading on the weight scores being inherently error-based since they are being optimized in regards to the error and constraints). Shahhosseini/Rendon fails to explicitly disclose but Elmachtoub discloses wherein the calculating the error-based weight scores comprises reflecting, by the processor, a first parameter value for adjusting a deviation between the error-based weight scores (Elmachtoub [Page 6 Paragraph 2]; PNG media_image9.png 571 658 media_image9.png Greyscale PNG media_image10.png 393 649 media_image10.png Greyscale Wherein the loss function represents a deviation between the error-based weight scores (cTw), and the loss function value is adjusted by the cost reflected by vector parameter c) It would have been obvious to modify Shahhosseini/Rendon’s weight optimization function to incorporate Elmachtoub’s explicit loss function representative of deviation along with a cost parameter to adjust loss function deviation. One would have been motivated to do so because “A natural loss function in this framework is to measure the suboptimality of the decisions induced by the predicted input parameters, as opposed to measuring loss using input parameter prediction error” (Elmachtoub [Page 1]) thus allowing Shahhosseini/Rendon’s optimization to additionally take into account suboptimality of the predicted input parameters instead of the input parameter prediction error alone. Response to Arguments The Examiner acknowledges the Applicant’s amendments to Claims 2, 4, 13 and 15 . Applicant’s arguments filed February 9 th , 2026 , traversing the rejection of claims 14-20 under 35 U.S.C. § 112 have been fully considered, and are fully persuasive. Applicant’s arguments filed February 9 th , 2026 , traversing the rejection of claims 1-20 under 35 U.S.C. § 101 have been fully considered, and are fully persuasive. Applicant’s arguments filed February 9 th , 2026 , traversing the rejection of claims 1-3, 5-20 under 35 U.S.C. § 102(a)(1) and claim 4 under 35 U.S.C. § 103 have been fully 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. Conclusion 07-40 AIA 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 JONATHAN J KIM whose telephone number is (571)272-0523. The examiner can normally be reached 8-6. 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, Matt Ell can be reached on (571) 270-3264. 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. /JONATHAN J KIM/Examiner, Art Unit 2141 /MATTHEW ELL/Supervisory Patent Examiner, Art Unit 2141 Application/Control Number: 18/057,080 Page 2 Art Unit: 2141 Application/Control Number: 18/057,080 Page 4 Art Unit: 2141 Application/Control Number: 18/057,080 Page 5 Art Unit: 2141 Application/Control Number: 18/057,080 Page 6 Art Unit: 2141 Application/Control Number: 18/057,080 Page 7 Art Unit: 2141 Application/Control Number: 18/057,080 Page 8 Art Unit: 2141 Application/Control Number: 18/057,080 Page 9 Art Unit: 2141 Application/Control Number: 18/057,080 Page 10 Art Unit: 2141 Application/Control Number: 18/057,080 Page 11 Art Unit: 2141 Application/Control Number: 18/057,080 Page 12 Art Unit: 2141 Application/Control Number: 18/057,080 Page 13 Art Unit: 2141 Application/Control Number: 18/057,080 Page 14 Art Unit: 2141 Application/Control Number: 18/057,080 Page 15 Art Unit: 2141 Application/Control Number: 18/057,080 Page 16 Art Unit: 2141 Application/Control Number: 18/057,080 Page 17 Art Unit: 2141 Application/Control Number: 18/057,080 Page 18 Art Unit: 2141
Prosecution Insights