Through the methods of correlation analysis and main factor analysis, the relationship between the poplar INA bacte-rial canker and circumstances was analyzed and 9 main factors for affecting the disease were selected...Through the methods of correlation analysis and main factor analysis, the relationship between the poplar INA bacte-rial canker and circumstances was analyzed and 9 main factors for affecting the disease were selected. Based on the compre-hensive analysis of main factors and induced factors, the standard for risk grades of this disease was promoted and northeast region of China was divided into 4 districts with different risk grades: seriously occurring district, commonly occurring district, occasionally occurring district, and un-occurring district. Nonlinear regression analysis for six model curves showed that the Richard growth model was suitable for describing the temporal dynamics of poplar INA bacterial canker. By stepwise variable selection method, the multi-variable linear regression forecasting equation was set up to predict the next year抯 disease index, and the GM (1,1) model was also set up by grey method to submit middle or long period forecast.展开更多
基金National Foundation of Ninth Five-Year Plan (No. 96-005-04-01-03).
文摘Through the methods of correlation analysis and main factor analysis, the relationship between the poplar INA bacte-rial canker and circumstances was analyzed and 9 main factors for affecting the disease were selected. Based on the compre-hensive analysis of main factors and induced factors, the standard for risk grades of this disease was promoted and northeast region of China was divided into 4 districts with different risk grades: seriously occurring district, commonly occurring district, occasionally occurring district, and un-occurring district. Nonlinear regression analysis for six model curves showed that the Richard growth model was suitable for describing the temporal dynamics of poplar INA bacterial canker. By stepwise variable selection method, the multi-variable linear regression forecasting equation was set up to predict the next year抯 disease index, and the GM (1,1) model was also set up by grey method to submit middle or long period forecast.
