DSAS Colloquium Talk - Jonathan Yu-Meng Li
Room: 248
Speaker: Jonathan Yu-Meng Li
(Telfer School of Management, University of Ottawa)
Title: Optimization for measuring risk in stochastic programs
Abstract:
Accounting for the adverse impact of "non-average" events has become essential in many applications involving decision making under uncertainty. Its implementation through decision models, namely stochastic programs, requires careful measurement of risk that reflects one's concern about uncertain outcomes. Important theories such as convex risk measures outline conditions required for risk measurement but provide little guidance for cases not meeting the conditions, which unfortunately more often than not is the case in real-life situations. In particular, in this talk, we study cases where the distribution required by a law invariant risk measure is not available and/or the risk preference required by a risk measure cannot be identified. We aim to provide theoretical, computational, and empirical evidence that in these cases optimization can be a powerful tool to measure risk in a systematic fashion that is hard to achieve otherwise. Applications to operation management and finance will be presented.