A study of stochastic evolutionary game prediction models based on SVM and random forests

This paper focuses on the application of four algorithms in data prediction. In the paper, the paper firstly establishes a SVM model and random forest prediction model to judge the performance effect of the model. By calculating the prediction accuracy of the models and observing the ROC curves and confusion matrix, it was found that the two models had the shortcomings of low sensitivity and biased precision, respectively. To address this problem, this paper innovatively complements the advantages and disadvantages of the two models through game theory, and establishes a new model based on the SVM model and the stochastic model. A stochastic evolutionary game model based on the SVM model and the random forest model is established, and the accuracy reaches 99.70%, with MSE=0.0027 and RMSE=0.0522. At the same time, this paper also tries to use the integrated classifier XGBoost, which has a better performance, and the prediction accuracy reaches 99.9%.