Description
Allen Holder
Professor, Department of Mathematics, Rose-Hulman Institute of Technology
“Classifying with DEA and Robust Optimization”
We extend a classic DEA model to account for uncertain data. The new uncertain DEA (uDEA) model advances the intent of traditional DEA by allowing a decision-making unit to select both its evaluative function and its data set, the latter of which is selected from a collection of uncertain possibilities. The uDEA framework recasts the traditional use of linear programming as the modern employ of robust optimization. This means that the optimization problems of uDEA are convex; however, the classification problem that uses uDEA suffers from both a loss of convexity and egregious combinatorial growth. We develop ways to overcome these challenges and present applications in classifying stocks, identifying peer groups within the airline industry and political redistricting.
Biography
Allen Holder is a mathematician at the Rose-Hulman Institute of Technology and a longstanding member of INFORMS. He won the 2000 Pierskalla Award, the 2015 Rose-Hulman Board of Trustees Outstanding Scholar Award, the 2019 ICS Distinguished Service Award, a 2017 Volunteer Service Award, and a 2013 Moving Spirit Award. He has also chaired the INFORMS Health Application Section, served on several conference committees, and held editorial positions within OR and optimization. His research largely combines optimization and computational science to solve problems in medicine, the life sciences, and healthcare, but he also dabbles in topics like personnel management, manufacturing, economics, and art. He has (co-)authored about fifty papers and one textbook on >computational science.
Allen Holder, Professor, Department of Mathematics, Rose-Hulman Institute of Technology
Location
BB 1.01.20L, Multipurpose Room
Category:
Campus Events
