Lecture
This lecture covers Two-Stage Stochastic Programs, focusing on problem reformulation and Benders decomposition algorithm. It explains how to fix x to determine optimal second stage decisions yw separately, and solve the master problem to optimize over x. The instructor discusses the characteristic structure of these programs and how to exploit it. The lecture also delves into primal and dual subproblems, extreme points, recession cones, and extreme directions. It concludes with global and local sensitivity analysis, exploring the impact of changes in c, b, and problem constraints on optimality. The general framework for Local Sensitivity Analysis is presented, emphasizing the conditions under which the current basis remains optimal.