Klasifikasi kualitas mutu susu pasteurisasi menggunakan metode klasifikasi k-Nearest Neighbor

Abstract

Milk is a product that is often consumed because it has many good benefits for the body. One of the benefits of milk is that it contains calcium, which is useful for the growth of bones and teeth. However, many consumers still choose dairy products based only on their appearance, even though milk quality is not based on its appearance but is found during milk processing. Pasteurization is the process of heating whole milk at a specific temperature and for a certain period to increase the milk's shelf life and maintain the milk's quality. Two types of pasteurization processes are Long Temperature Long Time and High Temperature Short Time. k-Nearest Neighbor or k-NN is a data mining method used to classify objects from the data, using distance calculations or Euclidean Distance to look for similarities between neighbors. So, based on this problem, milk quality classification is carried out. The research was carried out by classifying milk quality using the K-NN method. The public data used to classify is 1,059 pasteurized milk data, with seven regular or ordinary attributes and one special or class attribute. The K values ​​that will be used in this research are k=3, k=5, and k=7. From the analysis of calculation results using the k-NN method, the accuracy values ​​obtained are k=3 is 97%, k=5 is 89%, and k=7 is 87%.