The Ground Temperature Prediction Model Based on Boosting Algorithm

Traditional statistical methods are difficult to meet the needs of the demands of temperature prediction. To improve the predicted performance of Boosting algorithm, this study adopts an improved Random Forest algorithm for the feature selection of meteorological data and employs Random Search and Bayesian Optimization for the parameter optimization of the model. The experimental result indicates that the Gradient Boosting Decision Tree(GBDT) based on Bayesian Optimization demonstrates significant superiority in temperature prediction.