Project Grant 2246417

The National Science Foundation awarded $235,438 under the Computer and Information Science and Engineering federal grant program to develop interpretable and fair machine learning frameworks, algorithms, and methodologies. The two-year project grant aims to address concerns that machine learning algorithms can inadvertently amplify human biases by funding research to design scalable, data-driven methods with provable fairness guarantees. Specifically, the grant will support the development of new sub-data selection frameworks to enhance fairness in learning outcomes, robust solution schemes to mitigate overfitting effects, and approaches to ensure algorithmic fairness in multi-stage decision-making problems. The research is being conducted by the Georgia Tech Research Corporation from October 2022 through March 2025 to inform policymakers about tradeoffs between machine learning results and social equity, and provide guidelines to promote diversity and fairness across relevant application domains.