XAI-Powered Smart Agriculture Framework for Enhancing Food Productivity and Sustainability
Top Cited Papers
Open Access
- 6 November 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 168412-168427
Abstract
A vital component of maintaining the world’s expanding populace is farming. In agriculture field, factors such as soil quality, weather patterns, and crop yields are essential components of usual possessions that affect farming manufacture. Despite advancements, prevailing smart systems quiet struggle with handling big amounts in prediction claims, often facing difficulties in balancing prediction accuracy and learning efficiency. To ensure sustainable food production, integrating advanced machineries such as machine learning and artificial intelligence in agriculture is essential. This study proposes an explainable AI (XAI)-based smart agriculture system to provide holistic recommendation for precision farming aimed at improving productivity while reducing environmental impact. We compiled a comprehensive weather, soil, and crop dataset from official and verified sources in India. From this dataset, we extracted and optimized features using pre-trained architectures and enhanced barnacles mating optimization (EBMO) algorithm, addressing the high-measure mentality and computational complexity issues often encountered in agricultural data analysis. The selected features were analyzed to provide holistic recommendations for precision farming using baseline ML models such as support vector machine, random forest, neural network, and decision tree. Additionally, we integrate the XAI framework to offer interpretable recommendation for optimizes the agricultural practices. The study developed an XAI-based smart agriculture system that provides holistic recommendations for precision farming to boost productivity. By using a comprehensive dataset and optimizing features with the EBMO algorithm, the research achieved high accuracy in crop yield predictions, particularly with the XLNet+SVM model, which outperformed existing models across various crops. The integration of SHAP and LIME further enabled interpretable AI-driven insights, enhancing transparency in decision-making. The results demonstrated significant improvements in prediction accuracy, resource management, and sustainability, offering valuable contributions to global food security through smart agriculture. The ability to explain these AI decisions further supports the adoption of AI technologies in agriculture, fostering resilient agricultural systems that are essential for feeding the world’s growing population.Keywords
Funding Information
- Deanship of Graduate Studies and Scientific Research, Jazan University, Saudi Arabia (RG24-S0140)
This publication has 35 references indexed in Scilit:
- Predicting Potato Crop Yield with Machine Learning and Deep Learning for Sustainable AgriculturePotato Research, 2024
- Near surface camera informed agricultural land monitoring for climate smart agricultureClimate Smart Agriculture, 2024
- Foundation models in smart agriculture: Basics, opportunities, and challengesComputers and Electronics in Agriculture, 2024
- Adaption of smart applications in agriculture to enhance productionSmart Agricultural Technology, 2024
- SPAGRI-AI: Smart precision agriculture dataset of aerial images at different heights for crop and weed detection using super-resolutionAgricultural Systems, 2024
- Applications of Artificial Neural Networks and Machine Learning in Civil Engineering2024
- A comprehensive survey on IoT and AI based applications in different pre-harvest, during-harvest and post-harvest activities of smart agricultureComputers and Electronics in Agriculture, 2023
- Factors affecting the adoption of climate-smart agriculture technologies in rice farming systems in Mali, West AfricaSmart Agricultural Technology, 2023
- XAI-CNVMarker: Explainable AI-based copy number variant biomarker discovery for breast cancer subtypesBiomedical Signal Processing and Control, 2023
- FARMIT: continuous assessment of crop quality using machine learning and deep learning techniques for IoT-based smart farmingCluster Computing, 2022