CVaR, Uncertainty Set, and the Joint Safeguarding Constraint<BR>

<p>
Jie Sun <BR>
Department of Decision Sciences, National University of Singapore jsun@nus.edu.sg <BR>

<P>
Abstract:<BR>

The condition-value-at-risk (CVaR) could be used to approximate a single safeguarding constraint (i.e., chance constraint) in stochastic optimization. There is a natural connection between a computationally tractable approximation to a CVaR constraint and an optimization problem with respect to an uncertainty set. This idea is extended to handling the joint safeguarding constraint problem, resulting practical improvement upon the current approach.