MACHINE LEARNING-BASED CROP RECOMMENDATION SYSTEM

Abstract

Agriculture is one of the primary contributors to the global economy and food security. Selecting the appropriate crop based on environmental and soil conditions is a critical factor in maximizing agricultural productivity and ensuring sustainable farming practices. Conventional crop selection methods primarily depend on farmers' experience and historical cultivation practices, which may not always yield optimal results due to changing climatic conditions and soil variability. This paper proposes a Machine Learning-Based Crop Recommendation System that utilizes supervised learning algorithms to recommend the most suitable crop based on soil nutrients, climatic conditions, and environmental parameters. The proposed framework employs data preprocessing, feature engineering, model training, hyperparameter optimization, and comparative performance evaluation using multiple machine learning algorithms, including Decision Tree, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), XGBoost, LightGBM, and CatBoost. Experimental analysis demonstrates that the optimized XGBoost classifier achieves superior prediction performance with an accuracy of 99.42%, outperforming other classifiers. The proposed system provides an intelligent decision-support tool that assists farmers in selecting suitable crops, improving yield, reducing cultivation risks, and promoting sustainable agriculture