Linear optimization problems have linear constraints and an objective that may be linear, quadratic or a general nonlinear function. They are very much the most popular model used in optimal decision-making. Since methods for solving discrete problems are frequently based on the solution of continuous relaxations, solution methods for continuous linear optimization are of primary importance in computational optimization.
This talk will introduce the two principal approaches to solving linear programming (LP) problems: the simplex method and interior point methods, with the aim of identifying their major computational components. Techniques for these will be discussed. The computational consequences when these methods are extended to permit the solution of problems where the objective is a quadratic or general nonlinear function will be identified.
|PDF (with pauses)||N71_Pauses.pdf|
|PDF (without pauses)||N71_NoPauses.pdf|