A Multi-Source Data-Driven Hybrid Model for Predicting Crop Yield under Varying Climatic Conditions

View at publisher Summarise with VIRAT
Authors: Premkumar Duraisamy, S. E. R. Sacha Emile R., Harishh D V, Jones Martin J
Cited by: 0
References: 20

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Crop yield prediction is essential for enhancing food security and promoting sustainable agriculture, especially under varying climatic conditions. This work introduces a multi-source datadriven hybrid model that integrates structured soilweather parameters with unstructured satellite imagery to achieve superior prediction accuracy. Structured data comprising soil nutrients, rainfall, temperature, and humidity are analyzed using the XGBoost algorithm, known for its robustness in handling complex non-linear interactions and feature importance analysis. Simultaneously, Convolutional Neural Networks (CNNs) are employed to extract spatial and vegetation features from NDVI and RGB satellite images, capturing crop health variations across regions. Experimental datasets were obtained from FAO, MOSPI, and NASA Harvest (Radiant Earth MLHub). The hybrid XGBoost-CNN framework demonstrated outstanding predictive performance, achieving an overall accuracy of 96.5%, outperforming individual models such as Random Forest (94.1 %) and LSTM (93.7 %). This work highlights the effectiveness of combining multi-source agricultural data with hybrid AI techniques to enhance yield forecasting accuracy, strengthen climate resilience, and support informed decision-making in precision agriculture.

Fields of Study & Subjects

Smart Agriculture and AI Remote Sensing in Agriculture Climate change impacts on agriculture

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
1055–1060
Publication date
21 January 2026
Language
EN