Article’s

The agricultural industry is the key source of livelihoods for many people that live in rural areas of India; however, it has many issues which affect the efficiency of the industry through non-optimized crop yield production and non-use of data in decision-making. The reliance on experience in farming means that there are difficulties experienced by farmers who can be negatively affected financially due to rapid movements in climate. This article discusses how the intelligent Crop Recommender System was developed through the use of machine learning (ML) to provide accurate forecasts of yields and crops to grow. Through the implementation of two separate ML predictive modules, namely the Crop Yield Predictor and Crop Recommender that utilized the Random Forest algorithm in a Python and Flask environment, the Crop Yield Predictor achieved an accuracy rate of 96.84% for yield predictions and the Crop Recommender achieved an accuracy rate of 87.56% for recommendation based on analysis of the 94 375 records in the data set. The Crop Recommender System allows users to enter their regional parameters e.g. soil type, season, area etc., which will provide the user with actionable information on growing crops. Additionally, the Crop Recommender System contains a fertilizer timing module which enables farmers to optimally apply resources. The migration from manual experience to automated predictive analytics will therefore allow farmers to enhance productivity, reduce crop losses and thus increase food security in the ever-changing agricultural sector.