Data-Driven Approach for Crop Yield Prediction Using Machine Learning Techniques

Abstract: Agriculture plays a vital role in food security and economic development, making accurate crop yield prediction important for effective farm planning and resource management. This study proposes a data-driven web-based agricultural decision-support system that integrates crop yield prediction, crop recommendation, real-time weather information, plant disease detection, farm analytics, and government scheme assistance. Machine learning algorithms such as Random Forest, XGBoost, and Gradient Boosting are used for crop yield prediction, while a Random Forest classifier is used for crop recommendation based on soil and weather parameters. The system further provides agronomic explanations, spray dosage calculation, multilingual support, voice input, and WhatsApp-based information sharing to improve accessibility and usability for farmers. The proposed KrushiMitra platform combines predictive analysis and practical agricultural support features within a single farmer-friendly interface.

Keywords: Crop Yield Prediction, Machine Learning, Random Forest, Precision Agriculture, Agriculture Decision Support.