Comparative Analysis of Soil Microbial Communities Using K-Nearest Neighbors and Naive Bayes Classification Models

This study aims to compare the performance of K-Nearest Neighbors (KNN) and Naive Bayes algorithms in classifying soil microbial communities based on various soil properties. The objective was to determine which model offers higher accuracy and computational efficiency for microbial classification. The study utilized publicly available soil microbial datasets containing features like pH, temperature, and soil texture, along with microbial abundance data. After pre-processing, data was split into training and test sets, and both models were evaluated using metrics such as accuracy, precision, recall, and F1 score. Results showed that KNN outperformed Naive Bayes, achieving an accuracy of 85% compared to Naive Bayes' 80%. KNN provided better classification performance by capturing complex, non-linear relationships in the data, while Naive Bayes was faster, with lower computational costs. In conclusion, KNN is recommended for tasks where accuracy is critical, while Naive Bayes is suited for larger datasets requiring faster computation.