Year: 2026 | Month: March | Volume 71 | Issue 1
Comparative Time Series Forecasting of Onion Cultivation and Export in India Using Traditional and AI- Powered Hybrid Models
Sudeshna Roy
S. Vivek Menon
Sona Augustine and Banjul Bhattacharyya
DOI:10.46852/0424-2513.1.2026.15
Abstract:
Agricultural trend prediction serves as a cornerstone for strategic policy formulation, market equilibrium maintenance, and sustainable national development. This investigation examined annual datasets encompassing onion cultivation area and production spanning 1978-2023, alongside export statistics from 1987-2023, employing comprehensive time series modeling and predictive analytics. The research methodology incorporated Auto-Regressive Integrated Moving Average (ARIMA) frameworks combined with advanced machine learning methodologies, including Time Delay Neural Networks (TDNN), Long Short-Term Memory (LSTM) architectures, and Random Forest regression algorithms. Additionally, innovative hybrid frameworks were constructed to harness the complementary strengths of these approaches in capturing temporal dependencies and nonlinear data characteristics. Predictive performance evaluation utilized Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE) metrics to identify optimal forecasting methodologies. Results demonstrate that hybrid modeling approaches consistently deliver enhanced accuracy relative to individual techniques, establishing their efficacy in predicting onion area, production, and export dynamics. This investigation underscores the substantial benefits of integrating traditional statistical frameworks with state-of-the-art machine learning methodologies to address agricultural forecasting complexities.
Highlights
- Hybrid models outperformed individual models in prediction accuracy.
- The models also demonstrated strong capability in capturing temporal and nonlinear patterns.
- Highlights the effectiveness of integrating traditional and modern predictive methods for agricultural forecasting.
© This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited
Print This Article Email This Article to Your Friend
