Application of new neural network technology in traffic volume prediction

Abstract

In view of the disadvantages of traditional neural network technology application, neural network integration technology is applied to traffic forecast for the first time. Neural network integration is used to study the same problem with a finite number of neural networks, and the output of each network is synthesized, which significantly improves the generalization ability of the learning system.Based on Boosting and Bagging integration method, the neural network integration method is proposed based on divide and conquer strategy, and discussed the network weights allocation algorithm. Using these three kinds of neural network integration prediction model, the real-time traffic volume of a certain intersection in Zhengzhou city is predicted, and the result is better than that of single neural network forecasting method. The experiments show that the neural network integration is better used in traffic forecasting.