One approach to the design of learning systems is to extract heuristics from existing adaptive systems.Genetic algorithms are heuristic learning models based on principles drawn from natural evolution and selective breeding, Some features that distinguish genetic algr ithms from other search methods are:A population of structures that can be interpreted as candidate solutions to the given problem;The competitive selectaon of structures for reprd uction, based on each structure's fitness as a solution to the given problem; Idealized uenetzc o~er-ators that alter the selected structures-in order to create new structures for further testing.