Dataset Balancing Techniques and Supervised Learning Algorithms for Predictive Analysis of Rice and Corn Yields

Agriculture plays a fundamental role in the global socio-economic context, forming the pillar of communities and economies around the world. This study focuses on the impact of pesticides and fertilizers on the crop yield of rice and maize in India, employing supervised machine learning algorithms. The methodology begins with the understanding and pre-processing of an agricultural yield dataset, with a focus on dataset balancing to handle any imbalances in the target variable categories. Several data mining models are then implemented, and selected based on various performance measures. The final stage involves the prediction of crop yield in relation to pesticide and fertilizer levels, followed by the interpretation of the results. In summary, the work provides a starting point for the analysis of agricultural yield in varied geographic settings with diversity in climate and agricultural practices. The methodologies proposed in this study provide a useful framework for future research on agricultural yield optimization solutions.