PERBANDINGAN ALGORITMA K-MEANS DAN DBSCAN PADA CLUSTERING MAKANAN BERDASARKAN KANDUNGAN NUTRISI BERBASIS WEB

Food is the primary source of energy for the human body, but uncontrolled eating habits can increase the risk of various diseases, including obesity and cardiovascular conditions. Therefore, understanding the nutritional content of food is crucial to help individuals make healthier dietary choices. One approach to understanding the characteristics of food is clustering based on its nutritional content. This study aims to group foods based on their carbohydrate, calorie, protein, and fat content using the K-Means and DBSCAN algorithms. These algorithms were chosen for their ability to cluster numerical data without the need for categorical labels. However, one of the challenges in applying these algorithms is determining the optimal number of clusters. To address this, the Elbow Method, Davies-Bouldin Index, and Silhouette Score were used to evaluate and assess the quality of the clustering results. Based on the evaluation results, the K-Means algorithm achieved a Silhouette Score of 0.578 and a Davies-Bouldin Index (DBI) of 0.661. These values indicate that the clustering result is fairly good, although the separation between clusters is not fully optimal. In contrast, the DBSCAN algorithm showed better performance, with a Silhouette Score of 0.625 and a DBI of 0.328, indicating that the resulting clusters are more compact and well-separated. The results of the clustering process were then visualized through a web-based application developed using Streamlit, an open-source Python framework that allows for fast, lightweight, and efficient web interface development.
Keywords: Clustering, Food, Nutritional Content, K-Means, DBSCAN, Web