MGMT 101 Lecture Notes - Lecture 15: Efficient-Market Hypothesis, Linear Programming Relaxation, Integer Programming
Document Summary
Decision variables required to be integer or binary (0-1) variables. Rounding solution of regular linear program is not always appropriate. Lp relaxation: ignoring integer constraints for moment and treat problem as regular linear programming situation. The value of os of lp relaxation provides upper/lower bound depending on whether you want min/max for value of os of mixed linear program. Binary decision variables have two possible value: 0 or 1. Right-hand side =0 if corresponding binary decision variable is zero. K out of n constraint: d1+d2+d3=2 or d1+d2+d3<=2. Solution of lp relaxation provides upper bound of optimal value of of. Rounding solution of lp relaxation dn necessarily work. Nonlinear decr marginal cost as production levels incr. Value of stock option is nonlinear function of underlying stock price. In demand as price of product incr. Local optimum: solution that optimal in its immediate neighborhood. Global optimum: solution that is optimal overall. Index funds: designed to mimic performance of stock market as whole.
