Integer Optimization for Predictive & Prescriptive Analytics in High Stakes Domains

Explore integer optimization techniques for predictive and prescriptive analytics in high-stakes domains through this 50-minute conference talk presented by Phebe Vayanos from the University of Southern California. Delve into innovative discrete optimization models and algorithms designed to address critical issues in homeless services delivery, suicide prevention, and substance use prevention. Discover how to develop optimal interpretable, robust, and fair decision-tree models that assist in high-stakes decision-making processes. Examine the flexibility, theoretical benefits, and practical advantages of these approaches, which demonstrate significant improvements over existing methods. Gain insights from research papers on strong optimal classification trees, fair classification trees, prescriptive trees from observational data, and optimal robust classification trees. Enhance your understanding of cutting-edge applications of artificial intelligence and discrete optimization in solving real-world challenges.