Forecasting of mustard yield using ARIMA model under frequentist approach

Abstract

Forecasting helps the decision-makers to make their future decisions more precisely in agricultural field. The main objective of the paper is to forecast the Mustard Yield production in the district of Haryana through fitting of univariate Auto Regressive Integrated Moving Average (ARIMA) models using the time series data of 40 years For fitting the ARIMA models, a real data set is first examined for the presence of stationarity and is achieved by performing differencing After achieving stationarity, the most appropriate model is selected among the various competing models by using Akaike’s information criterion and Bayesian information criterion. The validity performance of developed models are compared by using Mean absolute percentage error (MAPE) and Root mean square error (RMSE). After experimenting with various lag values for these processes, an ARIMA (1, 0, 1) model was identified as effective for predicting crop yield.