查看文章

Neural networks have been used during several years to solve classification problems in the Artificial Intelligence. A neural network is inspired by the biological neuron of our human body and its performance depends directly on the design of the hidden layers, and in the calculation of the weights that connect the different nodes. On this paper, the structure of the hidden layer is not modified, as the interest lies only on the calculation of the weights of the system. In order to obtain a feasible result, the weights of the neural network are calculated or optimized by minimizing function cost or error. A Firefly Algorithm, which is an efficient but simple metaheuristic optimization technique inspired by natural motion of fireflies towards more light, is used for the training of neural network. The simulation results show that the computational efficiency of training process using Firefly Optimization technique.