Forecasting Method for Indian Food grains

Abstract

In this study an attempt is made to propose a method of forecasting technique for production of total food grains. In fact as a first step specify the linear and multiplicative regression models with available independent variables with and without one time period log dependent variable as one of the independent variables and select one of the above four specifications having maximum adjusted R2 for forecasting purpose. In the second step assume that each of the independent variable [Xi (i= 1,2…k) and Yt-1] in the model is related to time variable as linear, exponential and power functions. Estimate these relations for each of the independent variables and select the best one based on adjusted R2 and predict the future values for all independent variables. As a third step substitute the future values of the independent variables obtained from step 2 in the selected model obtained from step 1. The obtained values are called forecasted values through proposed method. As a fourth step we assume that the dependent variable Y is related to time as linear, exponential and power functions, select one of these estimated relations with maximum adjusted R2 for forecasting. Compare the forecasted values through proposed method and that of time series method using adjusted R2 or MSE criteria. This methodology is applied to India as a whole as well as six regions of India for forecasting total production of food grains assuming total production of food grains (rice, wheat, cereals and pulses) is a function of their corresponding irrigated area with and without one time period log dependent variable as one of the independent variables.