Data prediction method for improved PSO Based Neural Network Model

The present invention relates to computer engineering applications, the improved optimization algorithm is a Neural Network Prediction model data based on the particle group, performed by the following steps: Step 1: representation of data samples; Step 2: preprocessing the data; step 3: RBF neural network parameter initialization; step 4: PSO using two yuan center determines the number of neurons in the hidden layer and the hidden layer of the kernel function; step 5: initialize local each PSO algorithm parameter. Improved PSO based neural network model data prediction method can readily determine the number of hidden layer neuron RBF neural network, thereby improving the RBF neural network performance, improve accuracy of prediction data used in the present invention Meanwhile, in the present invention is based on the PSO improved neural network model with low complexity, robustness, and good scalability.