关键词: RBF神经网络, 遗传算法, 梯度下降算法, 用户规模预测, 混合算法

Abstract:

It is very important for the decision-making of  communication operators to accurately predict the scale of communication users. However, the existing conventional prediction methods have problems such as large prediction error, low prediction rate and so on. We study the user scale prediction model based on the RBF neural network, and in order to improve the prediction performance of the RBF neural network algorithm and enhance the convergence efficiency of the prediction model, we combine the gradient descent algorithm and the genetic algorithm to optimize the parameters of the RBF neural network. Example analysis shows that the hybrid RBF neural network prediction model is better than other traditional prediction models, and it has an advantage in predicting speed.

Key words: RBF neural network, genetic algorithm, gradient descent algorithm, user scale prediction, hybrid algorithm