GA-based Fractal and Neuroevolution Compressions

This paper presents a compression methodology using genetic algorithms and web-services to test it. The algorithm combines two well-known algorithms of fractal and neural network compressions, making some improvements in their implementation. Thus, in fractal compression we increased speed of compression by a genetic search and in neural networks we used neuroevolution approach for adaptive neural networks. This type of network uses previously processed data to increase the compression rate. In general case, our algorithm outperformed JPEG by 12%.