Multi-criteria Agriculture Recommendation System using Machine Learning for Crop and Fertilizers Prediction.

Abstract

Agriculture plays an essential role in the economies of developing countries such as India and contributes significantly to the gross domestic product (GDP). The escalation in population has led to an upsurge in food demand. Numerous challenges such as the selection of crops, fertilizers, and pesticides without considering the various parameters like types of soil, water requirement, temperature conditions, and profitability analysis of crops for a particular region may lead to degradation in the quality of crop, yield and profitability. With the advancement of Computational technologies, researchers are working on recommending crops according to soil condition, water requirement, and market profitability along with fertilizers recommendation, disease identification, and pesticide recommendation. Through this research, we propose a machine learning-based crop and fertilizer recommendation algorithm called AgriRec. We have utilized soil properties, water level, farm size, and minimum support price of crop and design a machine learning model which predicts crops for different seasons. Further, we propose another mechanism that processes the properties/details of soil, crop, and fertilizer to envisage a combination of fertilizer(s) for a given pair of soil and crop. Our algorithm is tested for 5000 land samples of Gujarat region with 24 different crop and it successfully recommends crops with 95.85% accuracy and fertilizer with 92.11% accuracy with 4 times better performance as compared to existing benchmark recommendation approaches.