Higher Accuracy of Crop Recommendation Using K-Nearest Neighbors Algorithm Algorithm Over Decision Tree Algorithm

The main goal of this study is to achieve higher accuracy in crop recommendation by evaluating and comparing the performance of different machine learning classifiers [1]. Specifically, we focused on two algorithms: the K-Nearest Neighbors (KNN) algorithm and the Decision Tree algorithm. We tested these algorithms on a dataset containing 2,201 records and conducted 10 iterations for each to assess their classification performance across various scales. The G-power test used for this experiment was set at 80%, ensuring reliable statistical power. The results showed that the K-Nearest Neighbors algorithm achieved a mean accuracy of 89.78% for crop recommendation, while the Decision Tree algorithm had an accuracy of 81.8%. An independent samples t-test revealed a statistically significant difference in accuracy between the two algorithms, with a significance value of 0.013. This study aims to introduce an innovative approach by utilizing recent machine learning classifiers for crop recommendation. The comparison indicates that the K-Nearest Neighbors algorithm outperforms the Decision Tree algorithm, making it a more effective choice for this application.