Enhancing Agricultural Decision-Making: A Supervised Machine Learning Approach

Agriculture in India is becoming an undesirable livelihood for many considering the uncertainty of making a profit due to either crop failure or an unfavorable market. The primary challenge Indian farmers face today is the lack of information such as weather patterns, ideal crop variety, and market trends. While many current systems try to address this problem, they fail to recognize the diversity in soil and environmental conditions in India, thus resulting in systems applicable in very niche scenarios. With nearly 70% of the Indian population engaged in agriculture, the first issue to be tackled is making an application widespread. Fortunately, smartphones and the internet exist in even the most remote regions of India. This application directly addresses the pressing issue by offering a decision support system that uses machine learning techniques to offer valuable information to farmers anywhere in India. The system has four unique utilities: Crop recommendation using soil quality or location, yield prediction, and price prediction. Various machine learning algorithms such as Decision Tree, Random Forest, KNN, and Naive Bayes were tested for each module to select the algorithm that provided the best accuracy. The machine learning models have been trained exclusively using Indian agricultural data from data.gov.in. The utility offered by the back-end ML modules is presented in a simplistic and intuitive user interface. With help of the information presented by the application, farmers can make informed decisions having a higher probability of success in terms of cultivation and profit.