PENERAPAN DATA MINING DENGAN METODE K-NEAREST NEIGHBOR (KNN) UNTUK MENGELOMPOKKAN MINAT KONSUMEN ASURANSI (PT.JASARAHARJA PUTERA)

Abstract

Insurance comes from the word insurance, which means insurance. Insurance is an agreement between the insured (customer) and the insurer (insurance company). Data Mining is a way of finding hidden information in a database and is part of the Knowledge Discovery in Databases (KDD) process to find useful information and patterns in data. K-Nearest Neighbor (KNN) is a method that uses supervised algorithms where the results of newly classified query instances are based on the majority of the label classes on KNN. The purpose of the KNN algorithm is to classify new objects based on attributes and training data. The KNN algorithm works based on the shortest distance from the query instance to training data to determine the KNN. One way to calculate the short distance or distance of a neighbor using the Euclidean distance method. Euclidean distance is often used to calculate distances. Euclidean distance functions to test the size that can be used as an interpretation of the proximity of the distance between two objects.Based on the results of the testing carried out this application is able to make predictions by looking at the smallest error value. In motor vehicle insurance the smallest average value is found at k = 4 at 0.103 and the highest accuracy value is at k = 2 by 42%, personal accident insurance the smallest average value is at k = 2 at 0.116 and the highest accuracy value is at k = 2 by 67%, and fire insurance the smallest average value is at k = 2 at 0.088 and the highest accuracy value is at k = 2 at 67%.Keywords—Data Mining, Insurance, K-Nearest Neighbor DOI: 10.5281/zenodo.3116132