基于CART 算法的空气质量指数回归预测模型的学习

摘要

In this paper , CART arithmetic and GCV are combined to establish AQI forecasting model.Experimental results show that the regression tree optimized by GCV can forecast AQI effectively.The regression tree has the best forecasting accuracy, compared with Multiple Linear Regression (MLR).%采用了CART算法结合广义交叉验证(GCV)，对空气质量指数(AQI)预测建立了模型。结果显示，采用广义交叉验证选出的最优回归树能有效的预测空气质量指数。与多元回归方法相比，回归树预测的准确率较高。