Please use this identifier to cite or link to this item:
|Title||Combine Genetic Algorithm and Particle Swarm Optimization Approach for Neural Network Classification|
|Title in Arabic||اسلوب دمج بيت الخوارزمية الجينية و خوارزمية السرب للتصنيف بطريقة الخلايا العصبية|
Artificial Neural Network (ANN) has played a significant role in many areas because of its ability to solve many complex problems that mathematical methods failed to solve. However, it has some shortcomings that lead it to stop working in some cases or decrease the result accuracy. This research proposed a new approach combining the most famous optimization algorithms, namely the particle swarm optimization algorithm (PSO) and the genetic algorithm (GA), to increase the classification accuracy of ANN. The proposed approach utilizes the advantages of both PSO and GA to overcome the local minima problem of ANN, which prevents ANN from improving the classification accuracy. It starts with finding out the best ANN using backpropagation algorithm through various attempts to use it as one of the population for the GA algorithm. If the solution is still not reached, PSO algorithm will start working with the half population which has the best fitness values. The process of keeping repeatedly applying GA followed by PSO with every time half of the last population with the best fitness values will be applied until the optimum solution is obtained. In contrary to other approaches, the proposed approach is domain independent, and has been evaluated by applying it using nine datasets with various domains and characteristics. The testing was performed with three main different approaches, first is only using the ANN without any optimization algorithms, the second is applying our proposed approach and the third is applying various methods presented in previous researches; GA alone, ANN followed by GA, PSO alone, ANN followed by PSO and GA followed by PSO. The comparison results show the superiority and the capability of our proposed approach for all the datasets to increase the classification accuracy whether the classification is high or low using other approaches.
|Publisher||الجامعة الإسلامية - غزة|
|Files in this item|