Evaluating Various Machine Learning Techniques for Crop Nutrients Prediction

Machine learning techniques assist very well in soil analysis and crop prediction. The technical analysis of soil and crop is the key factor in obtaining better yield and achieving sustainable agriculture goals. Ma-chine learning algorithms provide an efficient and well-framed structure for agriculture field analysis. This paper presents an approach to design various regression-based machine learning models to predict the required nutrients, rainfall, soil temperature, and humidity for a crop. The models were designed using a decision tree, linear regression, and random forest. The models were evaluated and their performance was compared. From experimental results, decision tree and linear regression showed comparative accuracy whereas, random forest showed the highest accuracy of 83.8 percent