文章基本信息

标题：An Implementation for Combining Neural Networks and Genetic Algorithms
本地全文：下载
作者：Md. Mijanur Rahman ; Tania Akter Setu
期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2015
卷号：6
期号：3
页码：218-222
语种：English
出版社：Ayushmaan Technologies
摘要：Neural Network (NN) and Genetic Algorithm (GA) are two very known methodology for optimizing and learning. Each having its own strengths and weakness. These two have generally evolved along separate paths. Recently there have been attempts to combine the two technologies. This research is devoted to implement a method for combining genetic algorithm with neural network (GANN). A collaborative approach has been used in this research. To integrated GA and NN into a single system, a population of neural networks is evolved, i.e., the goal of the proposed systemis to find the optimal neural network solution. In collaborative system GA and NN work parallel; for optimization it is very necessary to work with parallel.Several MATLAB functions and tools have been used to implement the proposed GANN method. The development and experiments demonstration of GANN is done on MATLAB 7.0.12. The proposed method consists of neural learning by the backpropagation algorithm and applies evolutionary, genetic operators. The learning behavior of the algorithm was tested on a simple pattern recognition example and it was able to prove its performance. The new ideas, concepts and processes of GANN bring new life in the field of Artificial Intelligence research.
关键词：Backpropagation;Collaborative Approach;Genetic Algorithm; Neural Network;Optimization