Enhancing Efficiency of Crop Recommendation using Incremental Rank-based Feature Selection Method

Traditional crop cultivation depends on the farmer’s choice, but nowadays, for suggesting suitable crops, a crop recommendation system is required for facilitating optimal crop yield predictions. This study investigates the effectiveness of an innovative approach, named as Incremental rank-based feature selection method, which can improvise the accuracy and efficiency of recommendation systems. Existing feature selection methods often stuck with the high-dimensional dataset as it is more complex. In contrast, proposed method systematically ranks and selects features based on their relevance to crop yield. Through rigorous experimentation and diverse agricultural datasets, it is demonstrated that the incremental approach consistently outperform than traditional methods in terms of predictive accuracy metrics such as Mean absolute error (MAE), Root mean squared error (RMSE), and R-squared ($\mathrm{R}^{2}$).Findings reveal that the incremental rank-based feature selection technique not only enhances predictive accuracy ($\mathrm{MAE}=0.245, \mathrm{RMSE}=0.313, \mathrm{R}^{2}=0.878$) but also significantly reduces computational complexity. The proposed method exhibits a runtime efficiency of 8.0 seconds, making it suitable for real-time decision-making. By iteratively selecting the most informative features, the method improves model interpretability and generalization capabilities.