Growth dynamics and forecasting of minor millets in India: A time series analysis

Abstract

The forecasting behaviour of millet plays a critical role in production planning at the Indian farm level. This study made an effort to forecast the area and production of small millets in India with time series analysis. The performance of the forecasting models was appraised and collated by the Mean Absolute Percentage Error (MAPE), Partial Autocorrelation Function (PACF) and Auto Correlation Function (ACF) criteria. For this analysis, the yearly data of the area and production of small millet from 1950 to 2021 were calculated. Among all Autoregressive Integrated Moving Average (ARIMA) models, ARIMA (0,1,0) was found to be the best fitted for forecasting the area and production of minor millets in India since, principally, this model relies on historical ideals of the sequences in addition to earlier error relations for forecasting minor millets and it does not adopt information of any fundamental model or associations as in some other approaches. The predicted values of minor millet area showed decreased trend from 422.4 thousand hectares in the year 2022 to 409.2 thousand hectares in the year 2026. Likewise, the production under small millets declined from 393.5 thousand tons to 159.5 thousand tons for the corresponding period. Hence, production of these crops can be enhanced by suitable use of inputs and timely application of inputs, high yielding varieties, government interventions like policy support, subsidising through the Public Distribution System and awareness by the way of propaganda and demonstration.