PENERAPAN ALGORITMA K-MEANS CLUSTERING UNTUK PENGELOMPOKAN KECELAKAAN BERKENDARA DI RUAS TOL JAKARTA-CIKAMPEK

Abstract

The high number of driving accidents on toll roads is an evaluation material. So that it becomes a special concern for toll road service providers, both state-owned and private, which have proven this concern by improving, adding infrastructure and educating road users to minimize accidents on toll roads. The initial stage of preventing driving accidents is to find out the factors that cause driving accidents obtained through accident data analysis. The analysis can be done with Data Mining, namely K-Means Clustering. K-Means Clustering groups the data into several clusters according to the characteristics of the data. The clustering stage is carried out by determining the number of cluster trials, namely by setting k = 3,
k = 5 and k = 7, and the performance is using the Davis Bouldin Index (DBI). The results of the cluster application of K-Means Clustering are tested to determine the best cluster model that tested and refers to the evaluation of DBI performance which approaches the value of Zero (Best value) sequentially so that the value of k=7 is the best DBI value of 0.179 while for k=5 the DBI is 0.180 and k=3 the DBI value is 0.233.