(1) Graduate School of Management, University of California, Irvine, CA 92697-3125, USA;(2) Decision and Information Sciences Department, Warrington College of Business, University of Florida, Gainesville, FL 32611-7150, USA
Abstract:
Conjugate function theory is used to develop dual programs for nonseparable convex programs involving the square root function. This function arises naturally in finance when one measures the risk of a portfolio by its variance–covariance matrix, in stochastic programming under chance constraints and in location theory.