A Hybrid Machine Learning Framework for Crop Yield Prediction Using Multivariate Agro-Climate and Soil Data

DOI

10.2991/978-94-6239-791-0_17How to use a DOI?

Keywords

Crop Yield Prediction; Machine Learning; Deep Learning; CNN-LSTM; Random Forest; Agriculture; Indian Dataset

Abstract

Accurate crop yield forecasting is essential for efficient planning of agricultural resources, food security and sound policy making. The paper proposes a hybrid deep learning framework for crop yield prediction imploying one-dimensional convolutional neural network and long short-term memory network. The model is trained on multivariate agro-climatic and soil data from and indian agriculture dataset convolutional techniques machine learning have been broadly used for this problem. However, they are not capable to capture the complex non-linear relationship that exist among climatic variables like temperature, rainfall, humidity and soil attributes. The proposed CNN-LSTM architecture overcome this limitation by jointly modeling static spatial features using the 1D-CNN module and sequential temporal climate pattern using the LSTM module and fusing the both representation to generate conclusive and definitive final yield prediction experimental evaluation on linear regression, decision tree, random forest and a stand-alone LSTM base line proposed model achieves at most prediction accuracy of 97.01% with the lowest MAE and RMSE among all compared models. The results indicate that the hybrid spatio-temporal deep learning model are more reliable and accurate for crop yield prediction as compared to the traditional machine learning methods, with practical value for farmers and agricultural planners.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Shrikant Karande
AU  - Rohan Jalan
AU  - Tejas Kedar
AU  - Vilas Ghonge
AU  - Amol Patil
PY  - 2026
DA  - 2026/10/09
TI  - A Hybrid Machine Learning Framework for Crop Yield Prediction Using Multivariate Agro-Climate and Soil Data
BT  - Proceedings of the International Conference on Artificial Intelligence and Machine Learning for Agro-Ecosystems (ICAIML-AE 2026)
PB  - Atlantis Press
SP  - 198
EP  - 207
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-791-0_17
DO  - 10.2991/978-94-6239-791-0_17
ID  - Karande2026
ER  -