DETAILED ACTION
This non-final office action is responsive to application 18/397,686 as submitted 27 Dec. 2023.
Claim status is currently pending and under examination for claims 1-20 of which independent claims are 1, 8 and 15.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged. However, the later-filed application must be an application for a patent for an invention which is also disclosed in the prior application (the parent or original nonprovisional application or provisional application). The disclosure of the invention in the parent application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994).
The disclosure of the prior-filed application, Application No. 63/477,547, fails to provide adequate support or enablement in the manner provided by 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph for one or more claims of this application. Particularly, independent claims 1, 8 and 15 recite “threshold influence” where neither term “influence” nor “threshold” is expressly supported by the provisional application. Therefore, claims have an effective filing date of 12/27/23, not the earlier filing of provisional 63/477,547.
Information Disclosure Statement
As required by MPEP 609(c), the applicant’s submissions of the Information Disclosure Statements dated 05/06/24 – 10/10/25 are acknowledged by the examiner and the cited references have been considered in the examination of the claims now pending. As required by MPEP 609 C(2), a copy of the PTOL-1449 initialed and dated by the examiner is attached to the instant office action.
Drawings
The drawings are objected to because: Figures 4, 6 and 8 are of insufficient quality such that the screenshots are too blurry to discern substantive detail. Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. The figure or figure number of an amended drawing should not be labeled as “amended.” If a drawing figure is to be canceled, the appropriate figure must be removed from the replacement sheet, and where necessary, the remaining figures must be renumbered and appropriate changes made to the brief description of the several views of the drawings for consistency. Additional replacement sheets may be necessary to show the renumbering of the remaining figures. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
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.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. In determining whether the claims are subject matter eligible, the examiner applies guidance set forth under MPEP 2106.
Step 1: Is the claim to a process, machine, manufacture, or composition of matter? Yes—all claims fall within one of the four statutory categories: claims 1-7 are a method/process, claims 8-14 are a computer-readable media/article of manufacture, and claims 15-20 are a system/machine. Thus, all claims are to statutory subject matter and the analysis should proceed per MPEP 2106.03.
Step 2A, prong one: Does the claim recite an abstract idea, law of nature or natural phenomenon? Yes—the claims, under the broadest reasonable interpretation, recites an abstract idea. In this case, claims fall within the enumerated grouping of abstract idea being “Mental Processes” and/or “Mathematical Calculations” but for the recitation of generic computer components. In particular, claims recite:
“generating explanatory data associated with a machine learning model” (Mental Judgment or Opinion, e.g. attributing biased output to subjective quality of an off-the-shelf model)
“calculating, based on the plurality of predictions, a predictive strength for the trained machine learning model” (Math Calculation or Mental Evaluation, e.g. post-hoc performance evaluation or suitability of an existing pretrained model)
“determining, based on the plurality of predictions and a plurality of features included in the trained machine learning model, one or more of the plurality of features having at least a threshold influence on the plurality of predictions” (Mental determination or a math calculation e.g. comparing min/max as greater- or less-than criteria for desired effect)
Focus of the claim concerns explanatory data affiliated with models based on calculating predictive strength and threshold influence. The terms ‘strength’ and ‘influence’ are abstract because they are not terms of art having established meaning, but broadly open the claim to subjective criteria without technical metrics. This could be performed by a human simply as a mental tally or template rubric for interpreting data generated from model result, or entail causality with counterfactuals. As such, the claims are drawn to at least mental processes and/or mathematical concepts as the abstract idea under MPEP 2106.04(a)(2).
Step 2A, prong two: Does the claim recite additional elements that integrate the judicial exception into a practical application? No—a practical application is not integrated by the judicial exception because the additional elements are as follows:
“computer-implemented” MPEP 2106.05(f) merely uses a computer as a tool to perform an abstract idea, e.g. [0115] “general purpose computer”
“receiving a trained machine learning model and a plurality of predictions generated by the machine learning model” MPEP 2106.05(g) adding insignificant extra-solution activity to the judicial exception, e.g. mere data gathering
“displaying, via a graphical user interface, one or more of the plurality of features and an indication of the predictive strength of the trained model” MPEP 2106.05(g) adding insignificant extra-solution activity to the judicial exception, e.g. necessary data outputting
Balance of the claim concerns computer-implementation with pre- and post-solution activities to receive and display. As set forth under MPEP 2106.04(a)(2) “A claim that requires a computer may still recite a mental process” such general-purpose computer is long found insufficient by the courts. Applying it to receive and display is merely input/output without distilling a concrete, real-world application or use-case. Therefore, the claim remains drawn to the abstract idea and additional elements fail to integrate the abstract idea into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No—the claims do not include additional elements that amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea in to a practical application, the additional elements are identified with respect to MPEP 2106.05 and do not demonstrate an inventive concept. Particularly, the additional elements are as follows:
“computer-implemented” MPEP 2106.05(f) merely uses a computer as a tool to perform an abstract idea, e.g. [0115] “general purpose computer”. Particularly, a general purpose computer does not qualify as a particular machine under MPEP 2106.05(b)
“receiving a trained machine learning model and a plurality of predictions generated by the machine learning model” MPEP 2106.05(g) adding insignificant extra-solution activity to the judicial exception, e.g. mere data gathering. Particularly, said extra-solution activity is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(i) “Receiving or transmitting data over a network, e.g., using the Internet to gather data”
“displaying, via a graphical user interface, one or more of the plurality of features and an indication of the predictive strength of the trained model” MPEP 2106.05(g) adding insignificant extra-solution activity to the judicial exception, e.g. necessary data outputting. Particularly, said extra-solution activity is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(iv) “Presenting offers” e.g. simple charting, plots, and/or supplemental evidence Romanowsky (below) PGP US2021/0390457A1 at [0082], Fig. 7.
Significantly more is not satisfied by the balance of the claim as the additional elements do not demonstrate a technical solution with particular transformation or meaningful limitation. If the claim language provides only a result-oriented solution, with insufficient detail for how a computer accomplishes it, then the claims do contain an inventive concept. As a whole, looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation.
For at least the foregoing reasons, claim 1 is found ineligible for patent. This rejection applies to independent claims 1, 8 and 15 as well to dependent claims 2-7, 9-14 and 16-20. Dependent claims when analyzed as a whole are held to be patent ineligible under 35 USC 101 because the additional recited limitations fail to establish that the claims are not directed to an abstract idea or that they include additional elements which integrate the judicial exception into a practical application or amount to significantly more.
Independent claim 8 discloses similar limitations to claim 1 and further recites “non-transitory computer-readable media storing instructions that, when executed by one or more processors” which is an additional element that falls under MPEP 2106.05(f) mere instructions to implement an abstract idea on a computer, and which particularly fails to satisfy test of particular machine under MPEP 2106.05(b). Therefore, the additional elements do not integrate the judicial exception into a practical application or amount to significantly more.
Independent claim 15 discloses similar limitations to claim 1 and further recites “A system comprising: one or more memories storing instructions; and one or more processors for executing the instructions” which is an additional element that falls under MPEP 2106.05(f) mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, and which particularly fails to satisfy the test of particular machine under MPEP 2106.05(b). Therefore, the additional elements do not integrate the judicial exception into a practical application or amount to significantly more.
Dependent claims 2, 9 and 16 disclose determining features for which increase in value increases predicted probability, and displaying features via graphical user interface. The limitation of determining is considered part of the abstract idea being mental determination and the limitation of displaying is an additional element which falls under MPEP 2106.05(g) insignificant extra-solution activity such as necessary data outputting and which is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(iv) “Presenting offers” e.g. simple charting, plots, and/or supplemental evidence Gathani (below) arXiv: 2109.06160v4 at Fig. 2. Therefore, the claim remains drawn to the abstract idea and the additional elements do not integrate the judicial exception into a practical application or amount to significantly more.
Dependent claims 3, 10 and 17 disclose determining features for which decrease in value decreases predicted probability, and displaying features via graphical user interface. The limitation of determining is considered part of the abstract idea being mental determination and the limitation of displaying is an additional element which falls under MPEP 2106.05(g) insignificant extra-solution activity such as necessary data outputting and which is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(iv) “Presenting offers” e.g. simple charting, plots, and/or supplemental evidence Gathani (below) arXiv: 2109.06160v4 at Fig. 2. Therefore, the claim remains drawn to the abstract idea and the additional elements do not integrate the judicial exception into a practical application or amount to significantly more.
Dependent claims 4, 12 and 19 disclose displaying via graphical user interface an indication of a number of features included in the trained machine learning model. The limitation is considered as additional elements which fall under MPEP 2106.05(g) insignificant extra-solution activity such as necessary data outputting and which is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(iv) “Presenting offers” e.g. simple charting, plots, and/or supplemental evidence Romanowksy (below) at [0072], Figs 6-7. No inventive concept is provided to meaningfully limit the claim. Accordingly, the additional elements fail to integrate the judicial exception into a practical application or amount to significantly more.
Dependent claims 5, 11 and 18 disclose receiving via graphical user interface an indication of a selected subset of predictions, calculating a quantity of predictions therein, and determining a comparative relationship. The calculating and determining limitations are considered part of the abstract idea being math calculations and mental determinations e.g. evaluative estimations. The limitation of receiving is considered as additional elements which falls under MPEP 2106.05(g) insignificant extra-solution activity such as mere data gathering which is a well-understood, routine and conventional activity under MPEP 2106.05(d)(II)(i) “Receiving or transmitting data over a network, e.g., using the Internet to gather data”. Therefore, the claim remains drawn to the abstract idea and the additional elements do not integrate the judicial exception into a practical application or amount to significantly more.
Dependent claims 6, 13 and 20 disclose calculating a quantitative lift value. The limitation is considered part of the abstract idea being math calculations which is enumerated as an abstract idea under MPEP 2106.05(a)(2). There are no additional elements.
Dependent claims 7 and 14 discloses wherein calculating comprises assigning qualitative labels to the model, the labels associates with a range of quantitative lift values. The limitation is considered part of the abstract idea embellishing calculations based on a range with assigned labels e.g. quantile or rank sorting. There are no additional elements.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 4, 6, 8, 12-13, 15 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over:
Doan Huu, Jacques, US PG Pub No 2023/0342659A1 hereinafter Doan Huu, in view of
Chen et al., “Explaining a series of models by propagating Shapley values” hereinafter Chen, and further in view of
Romanowsky et al., US PG Pub No 2021/0390457A1 hereinafter Romanowsky.
With respect to claim 1, Doan Huu teaches:
A computer-implemented method for generating explanatory data associated with a machine learning model {Doan Huu see [0033] “SHapley Additive exPlanations (SHAP values)” for a “value indicating an influence of that feature to the discriminated category which is generated by the trained secondary model” Figs 2-3 techniques, implemented by computer Figs 10-11}, the method comprising:
receiving a trained machine learning model and a plurality of predictions generated by the machine learning model {Doan Huu discloses [0061-62] “request a trained model… trained model to generate inferences” describes Fig 10 machine learning platform in communication with server, the inferences regard predictive estimates/likelihoods [0022-23], the trained model again at Fig 2};
determining, based on the plurality of predictions and a plurality of features included in the trained machine learning model, one or more of the plurality of features having at least a threshold influence on the plurality of predictions {Doan Huu discloses [0050] “determined discriminatory identifier features… features having an influence greater than a predefined threshold” so as for “features having a greatest influence on the output of the second trained model” model output being predictive estimates/likelihoods [0022-23] Fig 2}; and
However, Duan Huu does not appear to disclose a “predictive strength” which is taught by Chen:
calculating, based on the plurality of predictions, a predictive strength for the trained machine learning model {Interpreted per Instant Specification [0065] “predictive strength may be a quantitative value such as lift”. Chen teaches lift calculation at [P.11] Eqs. 6-8 “a ‘lift’ is defined… lift μ” and cont’d Eqs.10-15 applying the variable μ-lift. The predictions comprise an E-expectation function f(X) over Shapley features of a trained model e.g. Fig 7 “G-DeepSHAP… We train a model to predict” similarly at [P.5 Last¶], Fig 6};
Chen is directed to Shapley explanatory features for trained machine learning models thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to calculate lift/strength per Chen in combination for a motivation that [P.10 Last¶] “G-DeepSHAP works very well for a series of mixed model types” where ‘explicands’ are sampled from individual models [P.11 Eq.6] and may further help to “avoid bias” [P.4 ¶3]. See contributions and improvements over DeepLift [P.2 ¶3,8].
Chen further illustrates visualizations and plots throughout but does not expressly disclose “graphical user interface” which is disclosed by Romanowsky:
displaying, via a graphical user interface, one or more of the plurality of features and an indication of the predictive strength of the trained model {Romanowsky Fig 7:708-10, see [0082] “screenshot of a graphical user interface 700 used to display interaction Shapley values determined” cont’d “graphical user interface can include a suggested list 702 of feature interaction effects, sorted by strength of interaction” notes ‘strength’, e.g. [0058] “calculate the strength of feature interactions in a tree-based model”. See similar [0079] “display Shapley values” Fig 6}.
Romanowsky is directed to trained model interpretability/explainability with Shapley features thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to display via GUI the shapley features per Romanowsky in combination to arrive at the invention as claimed for a motivation [0057] “Advantageously… systems and methods described herein can be used to augment these visualizations to show Shapley values in a more complete context.”
With respect to claim 4, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 1, further comprising
displaying, via the graphical user interface, an indication of a number of features included in the trained machine learning model {Romanowsky [0072] “m is the number of features” Figs 6-7 show “Feature list” e.g. right of 608 or Fig 10:1004 GUI screenshots displaying Shapley features for a trained model like XGBoost [0058]. See similar [0055]}.
With respect to claim 6, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 1, wherein calculating the predictive strength for the trained machine learning model further comprises
calculating a quantitative lift value for the trained machine learning model {Chen [P.11] Eq. 6-8 e.g. “lift is the conditional expectation of the model’s output holding features” see Figs 6-7, 1-5}.
With respect to claim 8, the rejection of claim 1 is incorporated. The difference in scope being a non-transitory computer-readable media storing instructions executed by processor to perform limitations of method claim 1. Doan Huu discloses [0039] “non-transitory tangible medium” for “Software program code embodying these processes… executed by any one or more processing units…microprocessor” shown Fig 11. The remainder of this claim is rejected for the same rationale as claim 1.
With respect to claim 12, the combination of Doan Huu, Chen and Romanowsky teaches the one or more non-transitory computer-readable media of claim 8, and further teaches the limitation of claim 4. Therefore, the rejection of claim 4 is applied to claim 12.
With respect to claim 13, the combination of Doan Huu, Chen and Romanowsky teaches the one or more non-transitory computer-readable media of claim 8, and further teaches the limitation of claim 6. Therefore, the rejection of claim 6 is applied to claim 13.
With respect to claim 15, the rejection of claim 1 is incorporated. The difference in scope being a system comprising memory storing instructions executed by processor to perform limitations of method claim1. Doan Huu discloses [0066,68] “Hardware system 1100 may comprise a general-purpose computing apparatus and may execute program code… processor to execute program code” Fig 11 illustrates comprising memory 1160 and processing unit 1110 described e.g. [0039]. The remainder of the claim is rejected for the same rationale as claim 1.
With respect to claim 19, the combination of Doan Huu, Chen and Romanowsky teaches the system of claim 15, and further teaches the limitation of claim 4. Therefore, the rejection of claim 4 is applied to claim 19.
With respect to claim 20, the combination of Doan Huu, Chen and Romanowsky teaches the system of claim 15, and further teaches the limitation of claim 6. Therefore, the rejection of claim 6 is applied to claim 20.
Claims 2-3, 9-10 and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Doan Huu, Chen and Romanowsky in view of
Gathani et al., “Augmenting Decision Making via Interactive What-If Analysis” hereinafter Gathani (arXiv: 2109.06160v4).
With respect to claim 2, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 1, wherein each of predictions includes a predicted probability {Doan Huu [0038] “probabilities output by the trained second binary classification model” Fig 1}. Gathani teaches the computer-implemented method further comprising:
determining, based on the plurality of predictions and the plurality of features included in the trained machine learning model, one or more of the plurality of features for which an increase in a value of the feature increases the predicted probability included in a prediction of the plurality of predictions {Gathani [P.4 Rt.Col] “up-lift (positive, shown in green)… up-lift of 1.35%” similar at [P.5 ¶3] “up-lift of 48.65%” such that the up-lift is increase in predicted KPI features, “Every perturbation re-runs the model prediction to re-calculate the KPI value” using SystemD introduced [P.3 Sect.2] “We use Scikit-learn [27] to train machine learning models that predict KPI values… we verify the importances using traditional measures such as Shapley, Pearson, and Spearman rank… positive importance”}; and
displaying, via a graphical user interface, the one or more of the plurality of features {Gathani Fig 2 “user interface” screenshot at H and I showing 1.35% and 48.65% which are the up-lift indications described [P.4 Rt.Col], [P.5 ¶3]}.
Gathani is directed to what-if explanatory simulation with trained models thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to determine up-lift/increase for display per Gathani in combination for a motivation it [P.4 Last¶] “helps users to plan future actions that will help them achieve their KPI goals” describes goal inversion and sensitivity analysis, see [P.3 ¶4] “provide recommendations for changes needed in driver values to achieve user-specified KPI goals such as doubling the revenue or minimizing the churn rate”, note the contributions [P.3 Sect.1].
With respect to claim 3, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 1, wherein each of predictions includes a predicted probability {Doan Huu [0038] “probabilities output by the trained second binary classification model” Fig 1}. Gathani teaches the computer-implemented method further comprising:
determining, based on the plurality of predictions and the plurality of features included in the trained machine learning model, one or more of the plurality of features for which an increase in a value of the feature decreases the predicted probability included in a prediction of the plurality of predictions {Gathani [P.4 ¶6,1] “down-lift (negative, shown in red)” for “importance values range between -1 and 1 with extremes showing high negative” negative/down-lift is decrease in predicted KPI features, “Every perturbation re-runs the model prediction to re-calculate the KPI value” using SystemD introduced [P.3 Sect.2] “We use Scikit-learn [27] to train machine learning models that predict KPI values… we verify the importances using traditional measures such as Shapley, Pearson”}; and
displaying, via a graphical user interface, the one or more of the plurality of features {Gathani Fig 2 “user interface” screenshot at H and I showing lift indications described to comprise the down-lift [P.4 ¶6,1]}.
Gathani is directed to what-if explanatory simulation with trained models thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to determine down-lift/decrease for display per Gathani in combination for a motivation it [P.4 Last¶] “helps users to plan future actions that will help them achieve their KPI goals” goal inversion and sensitivity analysis, see [P.3 ¶4] “provide recommendations for changes needed in driver values to achieve user-specified KPI goals such as doubling the revenue or minimizing the churn rate”, note the contributions [P.3 Sect.1].
With respect to claim 9, the combination of Doan Huu, Chen and Romanowsky teaches the one or more non-transitory computer-readable media of claim 8, and further combination with Gathani teaches the limitation of claim 2. Therefore, the rejection of claim 2 with equal motivation is applied to claim 9.
With respect to claim 10, the combination of Doan Huu, Chen, and Romanowsky teaches the one or more non-transitory computer-readable media of claim 8, and further combination with Gathani teaches the limitation of claim 3. Therefore, the rejection of claim 3 with equal motivation is applied to claim 10.
With respect to claim 16, the combination of Doan Huu, Chen and Romanowsky teaches the system of claim 15, and further combination with Gathani teaches the limitations of claim 2. Therefore, the rejection of claim 2 with equal motivation is applied to claim 16.
With respect to claim 17, the combination of Doan Huu, Chen and Romanowsky teaches the system of claim 15, and further combination with Gathani teaches the limitations of claim 3. Therefore, the rejection of claim 3 with equal motivation is applied to claim 17.
Claims 5, 11, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Doan Huu, Chen, Romanowsky and Gathani in view of
Delgado et al., “The yield curve as a recession leading indicator: An application for Gradient boosting and Random Forest” hereinafter Delgado (arXiv: 2203.06648v1).
With respect to claim 5, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 3, further comprising
receiving, via the graphical user interface, an indication of a selected subset of the plurality of predictions {Romanowsky [0081] “For any selected feature, the graphical user interface 600 can show distribution plots for either the feature values and/or Shapley values… drop-down selector that allows a user to select any pair of features for display” for a “Shapley value in the prediction”. See also [0079-82] “displayed predictions” from a [0070] “subset S of input features”. Illustratively Figs 6, 10};
calculating, for the selected subset, a quantity of predictions included in the selected subset {Romanowsky [0086] “number of anomalous points or predictions (e.g., up to 500 predictions” and/or [0079] “distribution of predictions… 1st, 25th, 75th, and 99th percentiles” where the calculations comprise Eqs. 1-2 [0070,72]}; and
However, Romanowsky does not appear to disclose the following limitation which is met by Delgado:
determining a comparative relationship between first predicted probabilities associated with the selected subset and second predicted probabilities associated with the plurality of predictions {Delgado [P.6 ¶3] Eq.5 “predictions are compared… effect of withholding a feature depends on other features in the model, the preceding differences are computed for all possible subsets” a comparison may comprise [P.3] “correlation coefficient is used to verify collinearity… correlation analysis is shown between variables at Fig 2” similar at [P.8 ¶3] “By sorting the correlation coefficient, the most important variable is selected as the most correlated feature”. See also Lift Eq.1 [P.4 ¶4]}.
Delgado is directed to explanatory shapley features for trained models thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to compare predictions and correlate per Delgado in combination to arrive at the invention as claimed as applying known techniques to known methods ready for improvement to yield predictable results and/or a motivation [P.4 ¶9] “The advantage of this methods are that often provides predictive accuracy that cannot be beat, it can optimize on different loss functions and provides several hyperparameter tuning options that make the function fit” & [P.7 ¶2] “the most important part of this work comes with the feature importance as the first relevant output to interpret which variables are the main predictors.”
With respect to claim 11, the combination of Doan Huu, Chen, Romanowsky and Gathani teaches the one or more non-transitory computer-readable media of claim 10, and further combination with Delgado teaches the limitation of claim 5. Therefore, the rejection of claim 5 with equal motivation is applied to claim 11.
With respect to claim 18, the combination of Doan Huu, Chen, Romanowsky and Gathani teaches the system of claim 17, and further combination with Delgado teaches the limitation of claim 5. Therefore, the rejection of claim 5 with equal motivation is applied to claim 18.
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Doan Huu, Chen and Romanowsky in view of
M Y et al., US PG Pub No 2021/0142256A1 hereinafter M Y.
With respect to claim 7, the combination of Doan Huu, Chen and Romanowsky teaches the computer-implemented method of claim 6, wherein calculating the predictive strength for the trained machine learning model. M Y teaches further comprises:
assigning one of a plurality of qualitative labels to the machine learning model, wherein each of the plurality of qualitative labels is associated with a predetermined range of quantitative lift values {M Y [0056] “quality of the assigned label… assign the label b5” from {b1, b2, …, bk} for “a ‘lift’ metric with respect to the label” detailed Eq.1 “lift quantifies the relative propensity… higher lift” and a range may comprise e.g. “k>=5” and/or relatively higher lift as qualitative measure to distinguish the purchase labels from no-purchase labels [0055,57] as KPI of a marketing campaign, lift is probabilistic per [0022], see Fig 6 Lift column in the range of 1.21 – 1.95}.
M Y is directed to interpretability measures with trained models thus being analogous. A person having ordinary skill in the art would have considered it obvious prior to the effective filing date to assign labels using lift per M Y in combination to arrive at the invention as claimed for a motivation [0056] “intuitively, lift quantifies importance of the rule and corresponding user segment 120 (e.g., to a marketer, data analyst, and so on)…specify a useful target for future marketing campaigns” similar at [0022] “intuitively, the higher the lift, the increased importance of the rule to achievement of the KPI.”
With respect to claim 14, the combination of Doan Huu, Chen and Romanowsky teaches the one or more non-transitory computer-readable media of claim 13, and further combination with M Y teaches the limitation of claim 7. Therefore, the rejection of claim 7 with equal motivation is applied to claim 14.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Confortola et al., US PG Pub No 2023/0325692A1 replete with shapley influence threshold
Covert et al., “Explaining by Removing: A Unified Framework for Model Explanation” arXiv: 2011.14878v2 (co-author Lundberg, same as Chen) replete with feature removal
Jung et Oh, “Towards Better Explanations of Class Activation Mapping” arXiv: 2102.05228v2 discloses Lift-CAM, improvement over DeepLift
Dornadula et al., US PG Pub No 2022/0284499A1 generates shapley, influence threshold
Hisamitsu et al., US PG Pub No 2022/0327455A1 see Fig 18A:S303
Cheng et al., US PG Pub No 2022/0405623A1 Google explainability
Yang et al., US PG Pub No 2022/0114481A1 Salesforce counterfactual feature selection
Bierner et al., US PG Pub No 2023/0023202A1 assigns label to confidence interval (sim. lift)
Li et al., “InterpretDL: Explaining Deep Models in PaddlePaddle” Alg.1 import model
Agarwal et al., “openXAI: Towards a Transparent Evaluation of Post hoc Model Explanations” arXiv: 2206.11104v2 API interface
Yang et al., “OmniXAI: A Library for Explainable AI” arXiv: 2206.01612v8 see Figs 1, 3
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Chase P Hinckley whose telephone number is (571)272-7935. The examiner can normally be reached M-F 9:00 - 5:00.
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, Miranda M. Huang can be reached at 571-270-7092. 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.
/CHASE P. HINCKLEY/Examiner, Art Unit 2124
Prosecution Insights