Weather Forecast Model Based on Decision Tree and Neural Network

This paper is dedicated to weather prediction by establishing two models, the decision tree ID3 algorithm and the BP neural network algorithm, and analyzing the contingency and inevitability of summer wildfires in Chongqing. The decision tree ID3 algorithm uses information entropy for feature selection and builds a tree model. The BP neural network algorithm adjusts weights and biases through forward propagation and back propagation for prediction. First, the establishment steps of the two models are described, including information gain calculation and recursive construction of decision trees, as well as forward propagation and back propagation in BP neural networks. Secondly, this paper uses BP neural network to predict the weather in August. Finally, meteorological data were used to verify the model; the ID3 algorithm was used to analyze the contingency and inevitability of wildfires.