Abstract: Genetic programming combines and extends discrete decision theory with the principles of genetic and natural selection. The programs may be in the form of decision trees or diagram. The decision trees and diagrams are used in many discipline, genetic programming has many applications. Among those applications is pattern recognition. Different genetic programming techniques exist. This section describes a general technique for programs that use mathematical function. A function is routine that take one or two arguments, performs some function and returns value. The arguments with the routine are also functional routines, the resulting programs is like a tree in which each node represents a functional routine and each subtree an argument. Genetic programming with subtree crossover technique will be used that evolves a population over much iteration until some termination is satisfied. During each iteration the existing population is replaced by a new population that is derived from the existing population. The primary operations of reproduction and crossover are used for all problems. The two operations were sufficient for most of the problem to which the technique was applied in [9]

Keywords: Genetic Programming, Crossover, Decision Tree, Decision Diagram and Pattern Recognition