Forecasting and Regression Analysis for Sugarcane and Sugar Production in Karnataka

Abstract

Sugarcane is a major agricultural product in Karnataka as it is the third largest state to produce sugarcane and sugar in India. In this research we have collected the secondary data of 73 years (1950-51 to 2022-23) for the analysis in which only 69 years (1950-51 to 2018-19) had the data for sugar production.

The data that was collected was not stationary so it was made stationary by differencing by the order 2 and continued with the ARIMA model for forecasting. By observing the ACF and PACF plots different q and p values were selected and ARIMA model was prepared and finally ARIMA(2,2,1) was the best fit model. The best fit was confirmed by comparing the MAPE and AIC values of all the models that were prepared. Then ten years forecast was generated following the years 2022-23. By this people would be getting insight of the future of the industry and invest for profits.

Then to find if there is any linear relation between the sugar and area, sugarcane production, cane crushed and the number of sugar factories multiple linear regression model was built and was found that there is a linear relationship between the independent and the dependent variables. This was confirmed with the significant values that was provided as an output from the fit model analysis. The R-Square came out to be 99.93%, which tells us that the independent variables can explain 99.93% of the dependent variables from the model that has been found.