Evaluasi Performa Naïve Bayes dan CART pada Klasifikasi Kualitas Tahu

This Abstract has been read 30 times

Abstract

To remain competitive in the global market, tofu producers must ensure consistent product quality. Sumber Barokah Tofu Factory, a longstanding supplier of high-nutrient tofu, faces challenges in maintaining quality throughout the production process. This study compares the performance of Naïve Bayes and Classification and Regression Trees (CART) algorithms in classifying tofu quality using a dataset collected from a factory, which contains both high-quality and low-quality tofu samples. The research methodology encompasses problem identification, data collection, preprocessing, classification, validation, evaluation, and conclusion. Cross-validation was employed for model validation, and confusion matrices were utilised to assess precision, recall, and F1-score. Experimental results indicate that Naïve Bayes achieved an Accruracy of 91%, precision of 100%, recall of 85%, and F1-score of 92%, while CART achieved an Accruracy of 86%, precision of 70%, recall of 100%, and F1-score of 82%. These results suggest that Naïve Bayes is more suitable for classifying tofu quality in this context.

Keywords: Naïve Bayes, CART, Tofu Quality Classification, Cross-Validation