【摘要】 提出了非线性逐步回归、主分量非线性逐步回归和旋转主分量非线性逐步回归 3种因子选择方案 ,为提高褐飞虱长期预测的准确性提供依据。试验结果表明 ,主分量非线性逐步回归方法所选择的因子对褐飞虱的预测具有最好的效果 ,对独立样本的预测达到 80 %的正确率更多还原

【Abstract】 Three new methods for selection of the predictors are advanced as nonlinear stepwise regression,nonlinear stepwise regression with principal component and nonlinear stepwise regression with rotated principal component.The test results show that the forecasting model constructed with the predictors selected by nonlinear stepwise regression with principal component has the highest skill score of forecast for Brown Planthopper(BPH).The forecasting accuracy is 80 percent.更多还原