Machine Learning-Based Crop Recommendation System for Mizoram

Agriculture is an important sector of Mizoram domicile as more than half of its population relies on Agriculture as principle source of income and sustenance. Some farmers rely on the knowledge acquire from their parents through explicit explanation, and by observing and modelling their practices. But most farmer often struggle to understand the method and type of crops to cultivate for better crop yield. Even experienced farmer believed that using more fertilizer result in better crop yield in spite of that it damages the soil properties. To resolve this challenge, this paper presents a Crop Recommendation System using Machine Learning, tailored for Mizoram, enhancing Agriculture practices towards sustainable development. The Crop Recommendation System analyzes historical data, soil properties, weather pattern and crop performance to recommend the best crop for a specific region and its condition. The aim is to provide the Crop Recommendation System with information about the soil and the condition of the region. The study utilizes various Machine Learning Algorithms such as Random Forest, Decision Trees, Support Vector Machine and Logistic Regression to make optimal recommendation. The results indicate that Random Forest provides superior performance of 99% across all the evaluations metrics. Although many prevention measures need to be taken to avoid complications such as overfitting, etc. The overall results suggested that Random Forest achieved the best results as compared to all the other state of the art algorithms utilized with the same preprocessing steps. This approach enhances the crop and soil. After a long and often complicated process of farming method and selection of crop problem the Crop Recommendation System will aid Mizoram farmers to achieve better crops, yield and higher profit.