Epilepsy diagnosis using artificial neural network learned by PSO

摘要: Electroencephalogram (EEG) is used routinely for diagnosis of diseases occurring in the brain. It a very useful clinical tool classification epileptic seizures and epilepsy. In this study, epilepsy has been investigated using EEG records. For purpose, an artificial neural network (ANN), widely known as active technique, applied. The particle swarm optimization (PSO) method, which does not need gradient calculation, derivative information, or any solution differential equations, preferred training algorithm ANN. A PSO-based (PSONN) model diversified according to PSO versions, 7 models are described. Among these models, PSONN3 PSONN4 determined be appropriate due having better accuracy. methods-based versions compared with backpropagation algorithm, traditional method. addition, different numbers neurons, iterations/generations, sizes have considered tried. Results obtained from evaluated, interpreted, results earlier works done same dataset literature.