Open AccessAnalisis Cluster Ensemble dalam Pengelompokan Kabupaten/Kota di Provinsi Sulawesi Selatan Berdasarkan Indikator Kinerja Pembangunan Ekonomi Daerah
Author(s) -
Adrian Aqil Yusfar,
Muhammad Arif Tiro,
S. Sudarmin
Publication year - 2020
Publication title -
variansi
Language(s) - English
Resource type - Journals
ISSN - 2684-7590
DOI - 10.35580/variansiunm14626
Subject(s) - cluster (spacecraft) , hierarchy , group (periodic table) , similarity (geometry) , analytic hierarchy process , geography , computer science , data mining , statistics , mathematics , artificial intelligence , operations research , physics , political science , quantum mechanics , law , image (mathematics) , programming language
. Cluster analysis or group analysis is an analysis method to classify objects of observation into several groups based on their characteristics. Conventional methods namely Hierarchy and Non-Hierarchy are used in the formation of the initial group. However, the results of the grouping formed had mixed results so that the Cluster Ensemble analysis was then used to obtain a good final grouping. The Cluster Ensemble with the Link-Based Cluster Ensemble approach with the Connected Triple Based Similarity (CTS) method resulted in three final group divisions. The evaluation of the grouping performance used, namely Compactness and Davies-Bouldin, stated that the Cluster Ensemble was better than the hierarchical and non-hierarchical methods. The final group that has been formed is described using the average value for each variable in the district / city in South Sulawesi Province. The first group has the characteristics of regional economic development performance that is better than the second and third groups, but for the third group has the lowest characteristics of regional economic development performance from the first and second groups.Keywords : Cluster, Cluster Ensemble, Group Performance Evaluation, Performance, Regional Economic