This $954,310 project grant awarded by the National Science Foundation (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070) supports research and development to improve algorithmic decision-making systems. The project focuses on developing "responsive optimization" techniques to address issues of uncertainty, bias, and lack of transparency in algorithms used for high-stakes decisions in areas like college admissions, law enforcement, and personalized medicine. Key aspects include approximating the solution landscape, auditing solution robustness, ensuring explainability, and enabling interoperability with existing systems. The research aims to produce practical algorithms and general methods to enhance the trustworthiness of algorithmic decision-making. This 3-year project, awarded to Duke University, will also integrate the research into educational activities and cultivate a diverse group of young talent to tackle these important challenges.
Generated 2/25/25, 3:49 AM