Statistical modeling and forecasting of arima and arimax models for food grains production and net availability of India

Abstract

Agricultural development policies in India have aimed at reducing hunger, food insecurity, malnourishment and poverty at a rapid rate. The present work is designed with specific objectives to study the trend analysis of rice, wheat and total food grain in India for the period starting from 1950-2019. For stochastic trend model estimation, time series parametric regression models i.e. Linear model, Quadratic model, Exponential model, Logarithmic model, Auto Regressive Integrated Moving Average (ARIMA) and Auto Regressive Integrated Moving Average with explanatory variables (ARIMAX) were analyzed for estimating an appropriate econometric model to capture the trend of major food grain viz. rice, wheat, total food grain production and net availability of the country. Several goodness of fit criteria viz. Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), and maximum R-squared values was worked out for finding best fitted models. Kolmogorov-Smirnov (K-S) test and Run-test were used to estimate the „Normality‟ and „Independence‟ of residuals of all data series respectively. By using the best fitted models, it was observed that the availability of rice (70.05 kg/year), wheat (70.73 kg/year) and total food grain (182.96 kg/year) will decrease in 2021 as comparatively to this year.