Open Access Open Access  Restricted Access Subscription Access

Modelling and forecasting cotton production using tuned-support vector regression


Affiliations
1 Central Sericultural Research and Training Institute, Central Silk Board, Srirampura, Mysuru 570 008, India, India
2 ICAR-Indian Agricultural Statistics Research Institute, New Delhi 110 012, India, India
3 ICAR-Indian Institute of Rice Research, Hyderabad 500 030, India, India
4 ICAR-Indian Agricultural Research Institute, New Delhi 110 012, India, India
 

India is the largest producer of cotton in the world. For proper planning and designing of policies related to cotton, robust forecast of future production is utmost necessary. In this study, an effort has been made to model and forecast the cotton production of India using tuned-support vector regression (Tuned-SVR) model, and the importance of tuning has also been pointed out through this study. The Tuned-SVR performed better in both modelling and forecasting of cotton pro­duction compared to auto regressive integrated moving average and classical SVR models

Keywords

ARIMA, cotton production forecasting, SVR, time series, tuned-SVR.
User
Notifications
Font Size


  • Modelling and forecasting cotton production using tuned-support vector regression

Abstract Views: 472  |  PDF Views: 179

Authors

Amit Saha
Central Sericultural Research and Training Institute, Central Silk Board, Srirampura, Mysuru 570 008, India, India
K. N. Singh
ICAR-Indian Agricultural Statistics Research Institute, New Delhi 110 012, India, India
Mrinmoy Ray
ICAR-Indian Agricultural Statistics Research Institute, New Delhi 110 012, India, India
Santosha Rathod
ICAR-Indian Institute of Rice Research, Hyderabad 500 030, India, India
Sharani Choudhury
ICAR-Indian Agricultural Research Institute, New Delhi 110 012, India, India

Abstract


India is the largest producer of cotton in the world. For proper planning and designing of policies related to cotton, robust forecast of future production is utmost necessary. In this study, an effort has been made to model and forecast the cotton production of India using tuned-support vector regression (Tuned-SVR) model, and the importance of tuning has also been pointed out through this study. The Tuned-SVR performed better in both modelling and forecasting of cotton pro­duction compared to auto regressive integrated moving average and classical SVR models

Keywords


ARIMA, cotton production forecasting, SVR, time series, tuned-SVR.

References





DOI: https://doi.org/10.18520/cs%2Fv121%2Fi8%2F1090-1098