Genetic algorithms are a set of algorithms with properties which enable them to efficiently search large solution spaces where conventional statistical methodology is inappropriate. They have been used to find effective control and design strategies in industry, for finding rules relating factors and outcomes in medicine and business, and for solving problems ranging from function optimization to identification of patterns in data. They work using ideas from biology, specifically from population genetics, and are appealing because of their robustness in the presence of noise and their ability to cope with highly non-linear, multimodal and multivariate problems. This paper reviews the current literature on genetic algorithms. It looks at ways of defining genetic algorithms for various problems, and examples are introduced to illustrate their application in different contexts. It summarizes the different aspects which have been, and continue to be, the focus of research, and areas requiring further invetigation are identified.