【摘要】 为了提高年度台风频数预测准确率，利用中国气象局上海台风研究所提供的1951—2020年的台风样本数据和国家气候中心提供的142项环流特征量和海温指数资料，运用相关性分析方法筛选出高相关因子，再分别利用逐步回归方法和随机森林方法二次选取特征因子，确定最优特征子集，建立基于支持向量回归方法的年度台风频数预测模型，对比分析不同的特征选取方法对预测结果的影响。实验结果表明，融合智能方法选取特征的年度台风预测结果高于单一使用随机森林方法和逐步回归方法，支持向量回归方法平均绝对误差分别提高4.57%和4.90%。更多还原

【Abstract】 In order to improve the prediction accuracy of annual typhoon frequency, this paper uses the 1951—2020 typhoon sample data provided by the Shanghai Typhoon Research Institute of the China Meteorological Administration and 142 circulation feature quantities and SST data provided by the National Climate Center.Using the correlation analysis method to screen out high correlation factors, and then using the stepwise regression method and the random forest method to select the feature factors twice for the optimal feature subset. Establishing the annual typhoon frequency prediction model based on the support vector regression method, comparing and analyzing the influence of different feature selection methods on the prediction results. The experimental results show that the annual typhoon prediction results of the features selected by the integrated intelligent method are higher than those of the random forest method and the stepwise regression method alone, and the average absolute error of the support vector regression method is increased by 4.57% and 4.90%, respectively.更多还原