CLUSTERING WITH AGGLOMERATIVE HIERARCHICAL K-MEANS METHOD (CASE STUDY : RAIL ACCIDENT DATA IN EUROPE AND TRACER STUDY DATA IN ITB )

• Main Author: ISTI RIANI (NIM : 10113017), MILA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/23073
• Institution: Institut Teknologi Bandung

An analysis of Agglomerative Hierarchical K-Means is a technique of clustering data by combining Agglomerative Hierarchical and K-Means. Agglomerative Hierarchical used to establish the number of clusters and initial cluster centers. Agglomerative Hierarchical used is single linkage, complete linkage, and average linkage. K-Means used to group objects to the appropriate clusters. Agglomerative Hierarchical K-Means used Euclidean distance for clustering evaluation measures. Their application to rail accident data in Europe and Tracer Study data in ITB. For Rail Accident Data, the real life data set recorded by Annual Number of Accidents by Type of Accident at 31 countries in Europe, the clustering techniques process the data on six accident variables in 2014. And for Tracer Study ITB, the real life data set are analyzed in order group the study program Mathematics ITB alumnus form 2006 to 2009. The clustering techniques process the data on three variables, there are GPA, salary, and long waiting time get the first job. Clustering data with agglomerative hierarchical k-means has 1,5 time smaller cluster variance than using hierarchical only or k-means only.