Title: Study of Stationarity Concepts for a Class of Stochastic MPCC Problems.
Speaker: Arnab Sur, IIT Kanpur
Time and Date: 3:30 pm, Wednesday November 6, 2013
Venue: LCC 12, Lecture Hall Complex
Abstract: Mathematical programming problem with complementarity constraints (MPCC) is a special type of non-linear non-convex problem which fail to satisfy the well known constraint qualifications(like LICQ, MFCQ). Here we will discuss the stochastic version of the MPCC problems, i.e. objective and the constraint functions are random. As the well known constraint qualifications fail here so to develop the KKT condition is a real issue. First we will develop the KKT-type weak stationarity conditions and then show through examples that why we need to study stochastic MPCC problems separately by the help of weak stationarity conditions. Next we will discuss the strong and M-stationarity for the stochastic MPCC problems and perform the comparative study between the two stationarity concepts via sample average approximation (SAA) method.