Toll Free Helpline (India): 1800 1234 070

Rest of World: +91-9810852116

Free Publication Certificate

Vol. 12, Issue 2 (2023)

Development of hybrid time series models for forecasting autumn rice production in Assam

Author(s):
Borsha Neog, Bipin Gogoi, Smrita Barua, Supahi Mahanta and Kushal Sarmah
Abstract:
The study of forecasting in time series analysis has become a powerful tool in different applications in the agricultural field. The major goal of time series models comprises forecasting into the short- and long-term period, smoothening of irregular series and causal modelling of variables moving with time. Yearly data on production of Autumn rice have been used for forecasting from the year 1951 to 2018. The data from 1951-1998 were used for model building and 1999 - 2018 were used for checking the forecasting performance of the model. In this study ARIMA, ANN, SVM time series models and hybrid of both ARIMA-ANN, ARIMA-SVM were used to analyse the past behaviour of production of Autumn rice in order to make inferences about its future behaviour. ARIMA (2,1,2) model was selected as suitable model for Autumn rice and MAE for hybrid ARIMA (2, 1, 2)-ANN was found to be 34615.361 as compare to 39637.856 of ARIMA (2, 1, 2), MAE for hybrid ARIMA (2, 1, 2)-SVM was found to be 29464.313 as compare to 39637.856 of ARIMA (2, 1, 2) and 34615.361 of hybrid ARIMA-ANN. Hence, the performances of hybrid ARIMA-ANN and ARIMA-SVM were found to be better than that of ARIMA for both under training as well as testing data sets. And from the results, we found hybrid approach gives better results for forecasting of crop production.
Pages: 3376-3384  |  367 Views  211 Downloads


The Pharma Innovation Journal
How to cite this article:
Borsha Neog, Bipin Gogoi, Smrita Barua, Supahi Mahanta, Kushal Sarmah. Development of hybrid time series models for forecasting autumn rice production in Assam. Pharma Innovation 2023;12(2):3376-3384.

Call for book chapter