Prospects of livestock and dairy production in India under time series framework

Share of livestock sector especially dairy production in total gross domestic product (GDP) has shown a continuous rise trend over the last 30 years. Autoregressive integrated moving average (ARIMA) methodology was applied for modeling and forecasting of milk production of India. Auto-correlation (AC) and partial auto-correlation (PAC) functions were estimated, which led to the identification and construction of ARIMA models, suitable in explaining the time series and forecasting the future production. A significant increasing linear trend in the total milk production in India was found. To this end, evaluation of forecasting is carried out with mean absolute prediction error (MAPE), relative mean absolute prediction error (RMAPE) and root mean square error (RMSE). The best identified model for the data under consideration was used for out-of-sample forecasting up to 2015.