Optimizing Fertilizer Recommendations for Crops Using Gradient Boosted Decision Trees with Logistic Regression and GA

The research being conducted focusses on the essential issue of enhancing fertilizer recommendations for crops through the utilization of soil and environmental data. The proposed method combines Gradient Boosted Decision Trees with Logistic Regression, augmented by a Genetic Algorithm for hyperparameter optimization, to deliver accurate and efficient suggestions. In contrast to current methodologies, which frequently exhibit constrained accuracy and adaptability, our model shows substantial enhancement, attaining an accuracy of 95.4%, an F1 score of 94.8%, and a recall of 96.2%. The exceptional performance metrics demonstrate the model's efficacy in precisely forecasting appropriate fertilizer quantities and varieties under diverse situations. The study demonstrates the capacity of this innovative method to enhance precision agriculture by offering data-driven, real-time decision assistance for farmers, resulting in improved crop yields and resource management relative to conventional ways.