Association of Agrometeorologists

Forecasting models for predicting pod damage of pigeonpea in Varanasi region

PRITY KUMARI*, G.C.MISHRA and C.P. SRIVASTAVA

Present investigation considers comparison of time series statistical models like autoregressive integrated moving average (ARIMA) and artificial neural network (ANN) with explanatory multiple linear regression model for predicting per cent pod damage in pigeonpea by pod borer for Varanasi region of Uttar Pradesh using 27 years of data (1985-86 to 2011-12).The evaluation of best suited model was assessed by root mean squared error (RMSE). Based on empirical studies, ANN was found to be best suited model with lowest RMSE having forecasted per cent pod damage in pigeonpea by pod borer during the year 2012-13 for Varanasi region.

ANN ARIMA model, multiple regression, pigeonpea pod borer.