Prediction of milk yield using ARIMA

Abstract

Monthly milk yield data of Jersey and Holstein Friesian crossbred cows which were collected from Dairy Farm of Dr. Yashwant Singh Parmar University of Horticulture and Forestry Nauni from 1978 - 2014 to find out the most suitable forecasting method for milk production for sustainable future production and policy implications. ARIMA time- series (p, d, q) was applied to predict monthly milk yield over the years. The common approach modelling univariate time series is the autoregressive AR model. Autoregressive model is a linear regression of the current value of the series against one or more prior values of the series AR (p). The value of p is known as the order of the AR model. AR model has the straightforward interpretation. Another common approach for modelling univariate time series models is the moving average. A moving average is primarily a lagging indicator which makes it one of the most popular tools for technical analysis. Thirty seven years data on milk yield was used for modeling purpose. Moving Average ARIMA (6, 0, 2) was found the best fitted model for prediction of monthly milk yield.