Hybrid Mutation based Evolutionary approach for function optimization

Advent of Evolutionary algorithms (EA) is a major milestone in the field of data mining. Many research has been made to solve complicated mathematical and optimization problems since long. The Evolutionary Algorithm has been used effectively to resolve these optimization problems. Due to the evolutionary and stochastic nature of these algorithms, slow convergence rate is the major problem of these algorithms. We propose a new scheme to mutate the opposition Genetic Algorithm (GA). This technique is used to improve the population effectively by using the Gaussian Mutation (GM) and Cauchy Mutation (CM). Both the mutation schemes are used probabilistically. A suit of 5 optimization functions has been used to test the performance of the algorithm. The results are compared with Opposition based Genetic Algorithm (OGA) to evaluate the effectiveness of the presented algorithm. Proposed method shows results superior to GA and OGA for the majority of the test functions and shows comparable results over some functions.