G. Buvaanyaa and M. Gobi, Chikkanna Government Arts College, India

Crop suggestions are crucial because they help farmers select the best crops for their climate and region. In the past, this procedure mostly depended on specialized knowledge, which was labor-intensive and timeconsuming. Machine learning methods can be very useful in automating crop recommendations and detecting pests and diseases, allowing farmers to get the most out of their land while maintaining soil fertility and replenishing essential nutrients. The effectiveness of machine-learning algorithms for crop recommendation is investigated. To accurately predict which crops will be most suitable for a given location, the proposed method uses a variety of features, including soil nutrients and climate data. Crop suggestion could be revolutionized by this technology, which would benefit farmers of all sizes by increasing crop yields, sustainability, and overall profitability. By training and testing models with multiple configurations of machine learning algorithms, we have attained near-perfect accuracy through rigorous study of a massive historical data set. Across all models, the hybrid Algorithm achieves the best accuracy of 89.97%. This paper explores the theory, methodology, implementation, and evaluation of a hybrid RF-SVM classification model, highlighting its potential for improved accuracy and generalization.

Machine Learning, Feature Selection, Classification, Crop prediction, Hybrid Algorithm.