Characteristics and applications of random search methods for mathematical programming

In this paper we discuss the development of the random search method for mathematical programming problems. We present the characteristics of the random search approach for mathematical programming problems, and illustrate these characteristics in deriving a random search solution for a non-convex mathematical programming problem. We demonstrate its use in the ranking of the bases of a linear programming problem. A single machine scheduling situation is given and a random search method is employed in approximating a good solution. We also outline a real-life situation where an analytical solution may be difficult to obtain, but for which a random search may be useful.