Clustering in food waste analysis: Case study at student cafeteria

Abstract

Food waste is one of the most recent global issues that is getting more attention nowadays. Food waste is all food that is not consumed or thrown away even though it is fit for human consumption. This paper presents a method of direct weighing of edible food that is not consumed in the student canteen. Based on preliminary observations, food is classified into carbohydrates, vegetables, meat and others. The results of weighing data show more accurate results compared to previous study. With a total of 226 data, the research was focused on the carbohydrate and vegetable variables. The computer algorithm using the elbow method shows the unsupervised K-Means clustering of 4 clusters. The quality of the clustering method is good, with an average silhouette coefficient of 0.6. This research can be used as a basis for further studies on food waste and life cycle assessment.