Scenario decomposition of risk-averse multistage stochastic programming problems

We briefly discuss some history on the development of risk-averse optimization leading into coherent risk measures. For a riskaverse multistage stochastic optimization problem with a finite scenario tree, we introduce a new scenario decomposition method and prove its convergence. We then show how to apply our method to a typical operations management inventory and assembly problem. BIOGRAPHY Dr. Ricardo Collado is an assistant professor in the Howe School of Technology Management. He received his BS in Mathematics and Computer Science from the University of Puerto Rico and his Ph.D. in Operations Research from Rutgers University, New Jersey. Prior to joining Stevens in 2013, he was an Assistant Professor/Faculty Fellow at NYU’s Stern School of Business and a Professional Specialist at the Department of Operations Research and Financial Engineering from Princeton University. His research focuses on decision-making in the face of uncertainty and risk. Ricardo’s main research tools are Stochastic Optimization and Dynamic Programming, these coupled with a novel theory of Risk Measures. His main areas of applications are Energy and Financial Markets, and problems related to National Security. In addition, Ricardo has experience in Theoretical Computer Science, Graph Theory, Discrete Mathematics, and more classical mathematics such as Real and Complex Analysis, Topology, and Abstract Algebra. EVENT DETAILS Research Colloquium