Hybrid Deep Learning with Denoising Stacked Autoencoders for Agricultural Price Forecasting
Journal of the Indian Society of Agricultural Statistics · 30.07.2026 · 80 (01) · 149-163
DOI: 10.56093/jisas.v80i1.14 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the Indian Society of Agricultural Statistics, online erschienen am 30.07.2026. Verfasst von K.N. Singh und 6 weiteren Autor:innen.
Abstract (Originalsprache)
The complex nature of agricultural price data, characterized by perishability, seasonality, and non-stationarity, often renders conventional forecasting models inadequate. To address these challenges, this study proposes two hybrid learning frameworks: Stacked Autoencoder-based Deep Learning (SAE-DL) and Denoising Stacked Autoencoder-based Deep Learning (DSAE-DL). These models represent a significant departure from existing deep learning (DL) models, such as Multilayer Perceptron (MLP), Recurrent Neural Network (RNN), 1D Convolutional Neural Network (1D CNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). Unlike these standard models, which process raw input directly and often struggle with data volatility, our proposed hybrid approach integrates an autoencoder bottleneck to perform hierarchical feature extraction and dimensionality reduction prior to forecasting. The SAE-DL model focuses on capturing the underlying compressed representation of the price series, while the DSAE-DL introduces a stochastic denoising mechanism to specifically reconstruct core features from noisy data and outliers. Empirical analysis using a real-world onion price dataset reveals that the proposed DSAE-DL models, particularly the DSAE-RNN configuration, consistently outperform all other models across metrics, including RMSE, MAE, and MAPE. The Diebold-Mariano (DM) test further confirms that both proposed hybrid frameworks, SAE-DL and DSAE-DL, significantly improve prediction accuracy compared to standalone DL models. These findings highlight the superior robustness of hybrid autoencoder-based architectures in contending with market volatility.
Zitieren
APA
Singh, K., Avinash, G., Sharma, K., Kumar, R. R., Ray, M., Lama, A., & Shankar, S. V. (2026). Hybrid Deep Learning with Denoising Stacked Autoencoders for Agricultural Price Forecasting. Journal of the Indian Society of Agricultural Statistics, 80(01), 149-163. https://doi.org/10.56093/jisas.v80i1.14
Vancouver
Singh K, Avinash G, Sharma K, Kumar RR, Ray M, Lama A, et al. Hybrid Deep Learning with Denoising Stacked Autoencoders for Agricultural Price Forecasting. Journal of the Indian Society of Agricultural Statistics. 2026;80(01):149-163. doi:10.56093/jisas.v80i1.14
BibTeX
@article{Singh2026,
author = {Singh, K.N. and Avinash, G. and Sharma, Kamal and Kumar, Rajeev Ranjan and Ray, Mrinmoy and Lama, Achal and Shankar, S. Vishnu},
title = {Hybrid Deep Learning with Denoising Stacked Autoencoders for Agricultural Price Forecasting},
journal = {Journal of the Indian Society of Agricultural Statistics},
year = {2026},
volume = {80},
number = {01},
pages = {149--163},
doi = {10.56093/jisas.v80i1.14}
}
EUVIMED