Prosecution Insights
Last updated: November 18, 2026
Application No. 18/677,329

SYSTEM AND METHOD FOR MITIGATING BIAS IN CLASSIFICATION SCORES GENERATED BY MACHINE LEARNING MODELS

Non-Final OA §103§DOUBLEPATENT
Filed
May 29, 2024
Priority
Jun 03, 2020 — continuation of 12/002,258
Examiner
WONG, WILLIAM
Art Unit
2144
Tech Center
2100 — Computer Architecture & Software
Assignee
Capital One Financial Corporation
OA Round
1 (Non-Final)
32%
Grant Probability
At Risk
1-2
OA Rounds
59%
With Interview

Examiner Intelligence

Grants only 32% of cases
32%
Career Allowance Rate
132 granted / 414 resolved
-23.1% vs TC avg
Strong +27% interview lift
Without
With
+26.7%
Interview Lift
resolved cases with interview
Typical timeline
Avg Prosecution
19 currently pending
Career history
444
Total Applications
across all art units

Office Action

§103 §DOUBLEPATENT
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 in response to communications filed on 05/29/2024. Claims 1-20 are pending and have been examined. 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. Information Disclosure Statement The information disclosure statements submitted were filed on 06/26/2024, 06/26/2024, 06/26/2024, 08/09/2024, 02/11/2025, 04/01/2025, 09/08/2025, 12/30/2025, and 08/11/2026. The submissions are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner. Claim Objections Claims 4-6 and 15 are objected to because of the following informalities: As per claim 4, it appears that “the first and second sets of scores” in line 6 should be replaced with e.g. “the first set of scores and the second set of scores” for clarity and consistency. As per claim 5, it appears that “the trained machine learning model’s given set of input variables” in line 6 should be replaced with e.g. “a given set of input variables of the trained machine learning model” for clarity. This similarly applies to claims 6 and 15. Appropriate correction is required. 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. This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention. Claims 1-4, 7-8, 10-14, 16-17, and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Wei et al. (US 20210374581 A1) in view of Jiang et al. (“Wasserstein Fair Classification”, arXiv:1907.12059v1, Jul 28 2019, 15 pages cited in IDS dated 06/26/2024). As per independent claim 1, Wei teaches a computing platform comprising: at least one processor, at least one non-transitory computer-readable medium, and program instructions stored on the at least one non-transitory computer-readable medium that, when executed by the at least one processor (e.g. in paragraph 162, “at least one processor 16, coupled to the memory and the non-transitory computer readable medium, and operative to execute the instructions to be operative to perform any one, some, or all of the method steps”), cause the computing platform to: train a machine learning model by carrying out a machine learning process on a training data set (e.g. in paragraph 152, “existing training data 313” to “standard, non-fairness-constrained algorithm 317” to train the “existing classifier” as shown in figure 3 in contrast to being re-weighted to train the fairer classifier), wherein the trained machine learning model is configured to (i) receive an input comprising respective values for a given set of input variables (e.g. in paragraphs 55 and 148, “represent one or more protected attributes by a random variable A… X denote features used to predict Y… samples taken from two or more groups (e.g. protected group A and non-protected group B)”) and (ii) based on an evaluation of the received input, output a prediction of a given type (e.g. in paragraph 148, “obtaining, from an existing machine learning classifier 301, original probabilistic scores r(x)…into two or more classes… two classes, e.g. 1=“repay” and 0=“default”); detect bias in the trained machine learning model (e.g. in paragraphs 148 and 161, “aggregating the protected attributes or estimates thereof provided with each data point… either protected attribute information (what group each individual belongs to), or some estimate thereof, will be required”); after detecting the bias in the trained machine learning model, identify one or more input variable groups that contribute to the bias (e.g. in paragraph 161, “An example of an estimate is that a particular person is in Group A with probability 0.6. Information of either form permits inferring group distributions”); mitigate the bias in the trained machine learning model by producing a post-processed version of the trained machine learning model that comprises, for each respective input variable group of the identified one or more input variable groups, a respective transformation in place of the respective input variable group (e.g. in paragraphs 149-151, “transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the two or more groups” and figure 3); and use the post-processed version of the trained machine learning model to output a given prediction of the given type for a given input (e.g. in paragraphs 149-151, “as in FIG. 3 (post-processing mode), simply use the transformed probabilistic scores as classification output”), but does not specifically teach wherein input includes input vector. However, Jiang teaches input including an input vector (e.g. in pages 1-3, “xn is a feature vector that, possibly together with an, can be used to form a prediction ˆyn”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Wei to include the teachings of Jiang because one of ordinary skill in the art would have recognized the benefit of facilitating fair classification and/or representing input (also amounts a simple substitution that yields predictable results [e.g. see KSR Int'l Co v. Teleflex Inc., 550 US 398,82 USPQ2d 1385,1396 (U.S. 2007) and MPEP 2143(B)]). As per claim 2, the rejection of claim 1 is incorporated and the combination further teaches wherein the prediction of the given type comprises a score for use in rendering a classification decision (e.g. Wei, in paragraphs 149-151, “as in FIG. 3 (post-processing mode), simply use the transformed probabilistic scores as classification output”). As per claim 3, the rejection of claim 1 is incorporated and the combination further teaches input a first set of input vectors associated with an unprotected class into the trained machine learning model and thereby produce a first set of scores associated with the unprotected class (e.g. Wei, in paragraphs 55 and 151, “samples taken from…non-protected group B… transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the…groups”); input a second set of input vectors associated with a protected class into the trained machine learning model and thereby produce a second set of scores associated with the protected class (e.g. Wei, in paragraphs 55 and 151, “samples taken from…protected group A… transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the…groups”); perform a comparison between the first set of scores and the second set of scores and based on the comparison, determine that the trained machine learning model exhibits a threshold level of bias (e.g. Jiang, in pages 1-4, “raising concern that bias in the data and model inaccuracies can lead to decisions that are “unfair” towards underrepresented or historically discriminated groups… the threshold τ used… achieve SDP by enforcing the model output pdfs corresponding to groups of individuals with different sensitive attributes… use of the Wasserstein distance is motivated because this distance is defined and computable even between distributions with disjoint supports” and note formulas including Remark 1 on page 3). As per claim 4, the rejection of claim 3 is incorporated and the combination further teaches perform a comparison between distributions of the first and second sets of scores based on a Wasserstein distance metric (e.g. Jiang, in pages 1-4, “use of the Wasserstein distance is motivated because this distance is defined and computable even between distributions with disjoint supports” and note formulas including Remark 1 on page 3). As per claim 7, the rejection of claim 1 is incorporated and the combination further teaches wherein the respective transformation for at least one respective input variable group of the identified one or more input variable groups functions to at least partially neutralize the respective input variable group (e.g. Wei, in paragraphs 150 and 152, “transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the two or more groups… generate modified training data from the transformed probabilistic scores (re-weighting step 311 re-weights existing training data 313 to obtained re-weighted data 315”, i.e. neutralize). As per claim 8, the rejection of claim 1 is incorporated and the combination further teaches wherein the respective transformation for at least one respective input variable group of the identified one or more input variable groups functions to compress each input variable in the respective input variable group (e.g. Wei, in paragraphs 109, 150 and 152, “For the binary label-based metrics, we used the methods provided with the code for reductions… transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the two or more groups… generate modified training data from the transformed probabilistic scores (re-weighting step 311 re-weights existing training data 313 to obtained re-weighted data 315”; binary means compress). As per claim 10, the rejection of claim 1 is incorporated and the combination further teaches wherein, for each respective input variable group of the identified one or more input variable groups, the respective transformation utilized in place of the respective input variable group comprises a variable-specific transformation for each input variable in the respective input variable group (e.g. Wei, in paragraphs 55, 148, 150 and 152, “one or more protected attributes by a random variable A… groups (e.g. protected group A and non-protected group B)… transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the two or more groups… generate modified training data from the transformed probabilistic scores (re-weighting step 311 re-weights existing training data 313 to obtained re-weighted data 315”). As per claim 11, the rejection of claim 1 is incorporated and the combination further teaches wherein the bias comprises one or both of a positive bias component or a negative bias component (e.g. Wei, in paragraphs 134, 148, and 157, “groups (e.g. protected group A and non-protected group B)… fairness constraints can include, for example,…false positive parity, false negative parity”). Claim 12 is the medium claim corresponding to platform claim 1 and is rejected under the same reasons set forth. Claims 13-14, 16-17, and 19-20 are the method claims corresponding to platform claims 1, 3, 7-8, and 10-11 and are rejected under the same reasons set forth. Claims 5-6 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Wei et al. (US 20210374581 A1) in view of Jiang et al. (“Wasserstein Fair Classification”, arXiv:1907.12059v1, Jul 28 2019, 15 pages cited in IDS dated 06/26/2024) and further in view of Yang et al. (US 20210241033 A1 cited in IDS dated 06/26/2024) and Hickey et al. (US 20210248503 A1 cited in IDS dated 06/26/2024). As per claim 5, the rejection of claim 1 is incorporated and the combination further teaches wherein each respective input variable group includes a respective subset of the given set of input variables (e.g. Wei, in paragraphs 55, 148 and 161, “aggregating the protected attributes or estimates thereof provided with each data point… either protected attribute information (what group each individual belongs to), or some estimate thereof, will be required”), but does not specifically teach, as a whole, divide the trained machine learning model’s given set of input variables into a plurality of input variable groups; determine respective bias contribution values for the plurality of input variable groups using a score explainer function; and based on the respective bias contribution values that are determined for the plurality of input variable groups, identify the one or more input variable groups that contribute to the bias. However, Yang teaches divide a trained machine learning model’s given set of input variables into a plurality of input variable groups (e.g. in paragraphs 68-75, “target inputs are clustered into multiple clusters according to corresponding computed probabilities (or “propensities”)… Each cluster includes target inputs assigned to the sensitive group and other target inputs excluded from the sensitive group”), determine respective bias contribution values for the plurality of input variable groups using a score function and based on the respective bias contribution values that are determined for the plurality of input variable groups, identify the one or more input variable groups that contribute to the bias (e.g. in paragraphs 68-75, “determining whether the statistically significant difference exists between target inputs assigned to the sensitive group and other target inputs excluded from the sensitive group… label bias for the respective cluster is identified when the statistically significant difference is determined”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of the combination to include the teachings of Yang because one of ordinary skill in the art would have recognized the benefit of facilitating determination of bias, but does not specifically teach the score function including a score explainer function. However, Hickey teaches a score explainer function (e.g. in paragraph 209, “data points within training data 806 are assigned weight values by control unit 804 according to an estimated relationship between the corresponding predicted scores and the protected characteristic. The estimated relationship is based on an explicability score assigned to each data point. Preferably, the explicability score for a data point is a Shapley Additive Explanation (SHAP) value, which provides a unified methodology for interpreting model predictions”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of the combination to include the teachings of Hickey because one of ordinary skill in the art would have recognized the benefit of providing a unified methodology for interpretations (also amounts a simple substitution that yields predictable results [e.g. see KSR Int'l Co v. Teleflex Inc., 550 US 398,82 USPQ2d 1385,1396 (U.S. 2007) and MPEP 2143(B)]). As per claim 6, the rejection of claim 5 is incorporated and the combination further teaches divide the trained machine learning model’s given set of input variables into the plurality of input variable groups based on a clustering algorithm (e.g. Yang, in paragraphs 68-75, “target inputs are clustered into multiple clusters according to corresponding computed probabilities (or “propensities”)”). Claim 15 is the method claim corresponding to platform claim 5 and is rejected under the same reasons set forth. Claims 9 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Wei et al. (US 20210374581 A1) in view of Jiang et al. (“Wasserstein Fair Classification”, arXiv:1907.12059v1, Jul 28 2019, 15 pages cited in IDS dated 06/26/2024) and further in view of Garret et al. (US 20190384757 A1). As per claim 9, the rejection of claim 1 is incorporated and the combination further teaches wherein the respective transformation for at least one respective input variable group of the identified one or more input variable groups functions to compress a distribution of each input variable in the respective input variable group towards a value for the at least one input variable (e.g. Wei, in paragraphs 109, 150 and 152, “mean… evaluating the metric for each classifier in the distribution and then averaging weighted by their probabilities… transforming the original probabilistic scores into transformed probabilistic scores 309 that satisfy the fairness constraints 307 while minimizing loss in utility; e.g., with score transformer 303. The fairness constraints are with respect to the two or more groups… generate modified training data from the transformed probabilistic scores (re-weighting step 311 re-weights existing training data 313 to obtained re-weighted data 315”), but does not specifically teach a value including a median value. However, Garrett teaches a value including a median value (e.g. in paragraph 155, “taking the mean (across all M points, for each N) has a low-pass filtering effect (removing both high-frequency information and noise), while taking the median preserves high-frequency information in a non-linear fashion. The differing impact of the compression technique, mean vs. median, is accounted for by varying underling distributions”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of the combination to include the teachings of Garrett because one of ordinary skill in the art would have recognized the benefit of using well-known compression techniques (also amounts a simple substitution that yields predictable results [e.g. see KSR Int'l Co v. Teleflex Inc., 550 US 398,82 USPQ2d 1385,1396 (U.S. 2007) and MPEP 2143(B)]). Claim 18 is the method claim corresponding to platform claim 9 and is rejected under the same reasons set forth. 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 1-20 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-20 of U.S. Patent No. 12/002258. Although the claims at issue are not identical, they are not patentably distinct from each other because they recite similar subject matter which is obvious over one another. For example, claims 1-20 of ‘258 recite features of claims 1-20 in the present application, but does not specifically recite performing the comparison between distributions of the first and second sets of scores “based on a Wasserstein distance metric”. However, the prior art (e.g. Jiang as noted above) teaches perform a comparison between distributions of first and second sets of scores based on a Wasserstein distance metric (e.g. in pages 1-4, “use of the Wasserstein distance is motivated because this distance is defined and computable even between distributions with disjoint supports” and note formulas including Remark 1 on page 3). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the claims of ‘258 to include the teachings of the prior art (e.g. Jiang as noted above) because one of ordinary skill in the art would have recognized the benefit of facilitating fair classification. Conclusion The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. For example, Bhide et al. (US 11734585 B2) teaches “a post-processing computer-implemented method for post-hoc improvement of instance-level and group-level prediction metrics, the post-processing method including training a bias detector that learns to detect a sample that has an individual bias greater than a predetermined individual bias threshold value with constraints on a group bias, applying the bias detector on a run-time sample to select a biased sample in the run-time sample having a bias greater than the predetermined individual bias threshold bias value, and suggesting a de-biased prediction for the biased sample” (e.g. in column 2 lines 29-39). Any inquiry concerning this communication or earlier communications from the examiner should be directed to WILLIAM WONG whose telephone number is (571)270-1399. The examiner can normally be reached Monday-Friday 9am-5pm. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, TAMARA KYLE can be reached at (571)272-4241. 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. /W.W/Examiner, Art Unit 2144 09/29/2026 /TAMARA T KYLE/Supervisory Patent Examiner, Art Unit 2144
Read full office action

Prosecution Timeline

May 29, 2024
Application Filed
Oct 05, 2026
Non-Final Rejection mailed — §103, §DOUBLEPATENT (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12792932
Knowledge Graph Completion and Multi-Hop Reasoning in Knowledge Graphs at Scale
3y 8m to grant Granted Nov 10, 2026
Patent 12783940
TRANSFER KNOWLEDGE FROM AUXILIARY DATA FOR MORE INCLUSIVE MACHINE LEARNING MODELS
4y 1m to grant Granted Nov 03, 2026
Patent 12775857
PREDICTING MATCHING DENSITY WITH STRUCTURAL CAUSAL MODEL
4y 6m to grant Granted Oct 27, 2026
Patent 12736359
MACHINE LEARNING TECHNIQUES FOR TRAVERSAL PATH OPTIMIZATION
4y 4m to grant Granted Sep 15, 2026
Patent 12705009
SHARED AUGMENTED REALITY UNBOXING EXPERIENCE
2y 5m to grant Granted Aug 11, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

1-2
Expected OA Rounds
32%
Grant Probability
59%
With Interview (+26.7%)
Low
PTA Risk
Based on 414 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month