基于SVM的梅雨量预测方法

Prediction of Meiyu Based on Support Vector Machines

梅雨总量的预测研究对研究长江中下游夏季旱涝情况有重要的意义。根据近106a(1885-1990年)长江中下游沿江梅雨期的梅雨总量数据以及该时期梅雨特点,建立时间序列支持向量机(SVM)回归模型,并采用网格寻优参数函数对模型的参数进行优化,取得了较好的预测结果。

Prediction of Meiyu total is important in studying droughts and floods in the middle-lower reaches of Yangtze River in summer. According to recent 106a （1885-1990） data of the Meiyu in middle-lower reaches of Yangtze River and characteristics of the rainy period, built SVM regression time series model , and used parameter function of grid optimization to optimize the model parameters, and it has a better prediction result.