Discover Financial Services has patented a method, system, and computer-readable medium for detecting and mitigating bias in machine learning models. The process involves training the model, detecting bias based on input vectors, identifying biased input variable groups, and mitigating bias through post-processing. The patented technology aims to improve fairness in decision-making processes. GlobalData’s report on Discover Financial Services gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Discover Financial Services, UWB-enabled payments was a key innovation area identified from patents. Discover Financial Services's grant share as of May 2024 was 16%. Grant share is based on the ratio of number of grants to total number of patents.
Mitigating bias in machine learning models
A recently granted patent (Publication Number: US12002258B2) outlines a method carried out by a computing system to address bias in machine learning models. The method involves training a machine learning model on a training data set, detecting bias in the model relative to a protected class, identifying input variable groups contributing to the bias, and mitigating bias by producing a post-processed version of the model. The post-processed model applies transformations to the identified input variable groups to render classification decisions. The system uses the post-processed model to compare scores against classification thresholds, ensuring fair decision-making processes.
The patent also details the use of metrics such as Wasserstein distance, group partial dependence plot (PDP), and Shapley Additive Explanation (SHAP) to analyze bias contribution values and determine the effectiveness of bias mitigation techniques. The method separates bias into positive and negative categories, utilizes clustering algorithms for input variable grouping, and employs transformations to neutralize or compress input variables. Additionally, the patent highlights the application of the method in scenarios subject to review under the Equal Credit Opportunity Act (ECOA), emphasizing the importance of fair and unbiased classification decisions in sensitive domains. Overall, the patented method offers a comprehensive approach to identifying, analyzing, and mitigating bias in machine learning models, ensuring equitable outcomes in decision-making processes.
To know more about GlobalData’s detailed insights on Discover Financial Services, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
