Analyzing and Forecasting Cashew Production in Kerala using Time series and Regression Models

This paper deals with time series modelling approach (ARIMA Model) to forecast Cashew nut production in Kerala, India. The study assesses models related to area under cashew cultivation, production levels and productivity within Kerala. ARIMA (2, 0, 1) model was identified as the most suitable time series model for modelling the area of cashew cultivation in the state. Further, efforts were made to forecast, as accurate as possible, the Cashew nut production for a period up to ten years by fitting ARIMA(2,0,1) model to our time series data. The study was also explained different regression model techniques such as linear and multiple regression models, have been used to forecast cashew nut production. This paper also analyzes the combinations of linear and non linear regression models. For predicting cashew production with independent variable area, cubic model found to be best with highest R2 (0.96). For predicting cashew production with independent variable productivity, cubic model found to be best with highest R2 (0.59). Therefore Cubic model is the most appropriate model for predicting cashew production of the present study.