Comparative Performance Analysis of Decision Tree and K-Nearest Neighbors (KNN) Algorithms for Malformed Food Aroma Classification

One example of a significant food industry in Indonesia is the tofu industry. Tofu is very popular and has high nutritional value. However, the emergence of serious issues related to the use of hazardous materials such as formalin in food has become a serious threat to public health. Therefore, it is important to prevent the use of formalin in tofu production to protect the consumers’ health. This study aims to classify the aroma quality of tofu preserved with formalin based on its chemical components. The classification methods utilize k-nearest neighbors (kNN) and Decision Tree (DT). Evaluation of classification results using accuracy, precision, recall, and F1 score metrics. The experimental results with the Formalin Tofu Aroma dataset showed that KNN has 80.86% accuracy, 72.04% precision, 99.05% recall, and 83.44% F1-score. Meanwhile, the Decision Tree has 75.17% accuracy, 75.85% precision, 71.73% recall, and 73.70% F1-score. Based on these results, KNN is superior in terms of accuracy, recall, and F1 score. Whereas the Decision Tree outperforms in terms of precision.