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 .
Status of the Claims
Claims 1-20 are pending for examination.
Claims 1, 8 and 15 are independent Claims.
Claims 1-20 are rejected under 35 U.S.C. §102.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1-20 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Guha et al. (U.S. 2023/0350977 hereinafter Guha).
As Claim 1, Guha teaches a computer-implemented method, the method comprising:
maintaining a historical data sample comprising a plurality of observation records (Guha (¶0032 line 1-3), “subpopulation may represent a group of people who share a demographic attribute that is historically underrepresented and/or misrepresented”), each observation record comprising a set of predictive features, a set of dependent variables, and a baseline weight variable (Guha (¶0012 last 10 lines), “when presented with a female facial image as an input (predictive features), more than fifty percent of the output images (baseline weight variable) generated by the model depicted the subjects in a state of underdress relative to the output images generated based on male facial images (dependent variables) (which were more likely to depict the subjects wearing professional attire such as suits).”);
binning the plurality of observation records across the set of predictive features and the set of dependent variables to generate predictive feature bins and dependent variable bins (Guha (¶0033), “sample counts of the training data may indicate that a number of samples corresponding to the subpopulation (predictive feature bins) fails to meet some threshold relative to the number of samples for other subpopulations and/or a total number of samples in the training data (dependent variable bins), where the threshold may be configured by the party for whom the machine learning model is being constructed”),
generating a target distribution for the plurality of observation records (Guha (¶0034 line 1-5), “an insufficiency in the representation of the subpopulation may be identified automatically by analysis of the machine learning output (i.e., the output of the machine learning model), if the machine learning model has been constructed”), wherein the target distribution comprises a first plurality of target bin percentages for a subset of the predictive feature bins and a second plurality of target bin percentages for a subset of the dependent variable bins (Guha (¶0033 last 5 lines), “party may request that the number of samples for the subpopulation constitute at least x percent of the total number of samples in the training data, or constitute no less than y percent of the number of samples for any other subpopulation in the training data”);
generating a reweight variable for each of the observation records based at least in part on the target distribution and the baseline weight variable (Guha (¶0037 line 1-2, fig. 2 item 206), “the processing system may generate simulated data to mitigate the insufficient representation”), and
generating a reweighted data sample by replacing the baseline weight variable in each observation record of the historical data sample with a corresponding reweight variable (Guha (¶0038 line 11-14), “but in which the value for the feature pertaining to the insufficiently represented subpopulation (e.g., gender) has been modified to balance the representation.”).
As Claim 2, besides Claim 1, Guha teaches wherein each of the observation records further comprises a set of non-predictive variables, and wherein the set of non-predicative variables is binned to generate non-predicative variable bins (Guha (¶0032 line 6-8), “underrepresentation of the subpopulation in the training data may manifest itself in a lack of data points for accurate representation and/or comparison to the population at large”, underrepresentation is construed as non-predictive while overrepresentation is construed as variables).
As Claim 3, besides Claim 1, Guha teaches further comprising feeding the reweighted data sample to a classifier and training the classifier by the reweighted data sample (Guha (¶0034 line 1-5), “an insufficiency in the representation of the subpopulation may be identified automatically by analysis of the machine learning output (i.e., the output of the machine learning model), if the machine learning model has been constructed”).
As Claim 4, besides Claim 1, Guha teaches further comprising assessing a performance of a pre-existing machine learning model by the reweighted data sample (Guha (¶0034 last 8 lines), “For instance, the party may request that the model accuracy metric for the subpopulation be at least x, or the disparate impact score for the subpopulation be at least y. In one example, an insufficiency in the representation of the subpopulation may be revealed as a result of a test scenario that is input, along with the training data, into the machine learning model”).
As Claim 5, besides Claim 1, Guha teaches wherein generating the reweight variable comprises optimizing an objective function by minimizing a difference between the reweight variable and the baseline weight variable (Guha (¶0034 last 8 lines), “For instance, the party may request that the model accuracy metric for the subpopulation be at least x, or the disparate impact score for the subpopulation be at least y. In one example, an insufficiency in the representation of the subpopulation may be revealed as a result of a test scenario that is input, along with the training data, into the machine learning model”).
As Claim 6, besides Claim 5, Guha teaches wherein the optimizing the objective function is conditioned on a deviation between the target distribution and an achieved distribution does not exceed a pre-defined tolerance (Guha (¶0034 last 8 lines), “For instance, the party may request that the model accuracy metric for the subpopulation be at least x, or the disparate impact score for the subpopulation be at least y. In one example, an insufficiency in the representation of the subpopulation may be revealed as a result of a test scenario that is input, along with the training data, into the machine learning model”).
As Claim 7, besides Claim 1, Guha teaches wherein generating the reweight variable comprises optimizing an objective function by minimizing deviations between the target bin percentages and corresponding resultant bin percentages in the reweighted data sample (Guha (¶0034 last 8 lines), “For instance, the party may request that the model accuracy metric for the subpopulation be at least x, or the disparate impact score for the subpopulation be at least y. In one example, an insufficiency in the representation of the subpopulation may be revealed as a result of a test scenario that is input, along with the training data, into the machine learning model”).
As Claim 8, Guha teaches a system, comprising at least one programmable processor (Guha (¶0021), processor); and a non-transient machine-readable medium storing instructions that, when executed by the processor, cause the at least one programmable processor to perform operations ((Guha (¶0021), memory) comprising:
The rest of the Claim is rejected for the same reasons as Claim 1.
As Claims 9-14, the Claims are rejected for the same reasons as Claims 2-7, respectively.
As Claim 15- 20, the Claims are rejected for the same reasons as Claims 1-6, respectively.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Ahuja et al. (U.S. 2022/0180254) teahces reweighting sample based on domain changes.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NHAT HUY T NGUYEN whose telephone number is (571)270-7333. The examiner can normally be reached M-F: 12:00-8:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached at 571-270-5871. 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.
/NHAT HUY T NGUYEN/Primary Examiner, Art Unit 2147
Prosecution Insights