Vol. 17 No. 1
Estimation of incidence of pest and disease and pollulation of natural enemy of rice using weather parameters
Author(s): S.T. P. KUMAR AND D. MAZUMDAR
Abstract: In the present study, pest and disease incidence and population of damselfly in rice crop was estimated using conventional regression (Stepwise Regression) and Fuzzy linear regression. The weather parameters viz., Maximum Temperature, Minimum Temperature, Relative Humidity Morning, Relative Humidity Evening, Rainfall and Sunshine Hours were utilized as an explanatory variables (X’s) to build a prediction model. The performance of two models was evaluated on the basis of indicators such as Root Mean Absolute Error, Root Mean Square Error and average width of the prediction interval. It was found the average width of the prediction interval obtained for fuzzy linear regression was less compared to conventional stepwise regression analysis. In case of fuzzy linear regression, the prediction interval i.e., both upper and lower interval were close to the observed incidence due to less standard error of estimate (β). Fuzzy linear regression outperformed over conventional linear regression in predicting the incidence of pest and disease, and population of damselfly.
Download Full Article: Click Here