Pendekatan Modified Particle Swarm Optimization dan Artificial Bee Colony pada Fuzzy Geographically Weighted Clustering (Studi Kasus pada Faktor Stunting Balita di Provinsi Jawa Timur)

Hadi, Bambang Sulistyo (2017) Pendekatan Modified Particle Swarm Optimization dan Artificial Bee Colony pada Fuzzy Geographically Weighted Clustering (Studi Kasus pada Faktor Stunting Balita di Provinsi Jawa Timur). Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
1315201720 Master Tesis.pdf - Published Version

Download (4MB) | Preview

Abstract

Fuzzy Geographically Weighted Clustering (FGWC) adalah varian dari Fuzzy C-Mean (FCM), merupakan altenatif yang geographically aware untuk algoritma standar FCM dengan mendukung kemampuan untuk menerapkan efek populasi dan jarak untuk menganalisis cluster geo-demografis. FGWC sensitif terhadap inisialisasi ketika pemilihan pusat cluster secara acak menyebabkan solusi jatuh ke lokal optimum dengan mudah. Artificial Bee Colony (ABC) dan Particle Swarm Optimization (PSO) adalah metode yang cukup sering digunakan di metaheuristik. PSO dan ABC dapat memecahkan secara efisien dan efektif berbagai masalah fungsi optimasi dalam beberapa kasus ketika diintegrasikan ke dalam FCM. Pada penelitian ini akan di integrasikan PSO dan ABC kedalam FGWC (FGWC-PSO dan FGWC-ABC) dengan terlebih dahulu melakukan pemilihan penimbang inersia pada PSO dan modifikasi formula pada ABC sehingga diharapkan dapat meningkatkan performa FGWC. Selanjutnya kedua metode tersebut diterapkan untuk mengelompokkan kabupaten/kota berdasarkan faktor stunting balita karena stunting berkaitan dengan pola perilaku dan lokasi tempat tinggal (geografi), Stunting merupakan pertumbuhan linear yang terhambat dikarenakan kekurangan gizi pada masa penting pertumbuhan balita. Performa hasil cluster yang terbentuk akan dibandingkan dengan 6 Indeks evaluasi pengelompokkan yaitu Partition Coefficient, Classification Entropy, Partition Index, Separation Index, Xie and Beny Indeks, dan IFV Index.============================================================================================= Fuzzy Geographically Weighted Clustering (FGWC) is a Fuzzy c-mean (FCM) variant, which geographically aware alternative to a standard FCM algorithm by supporting the capability to apply population and distance effects for analyzing a geo-demographic cluster. FGWC is sensitive to initialization when the random selection in the cluster falling into the local optima easily. Artificial Bee Colony (ABC) and Particle Swarm Optimization (PSO) are the most popular methods on metaheuristic. PSO and ABC can solve efficiently and effectively various functions optimization problems, in some case when integrated into FCM. In this study, the PSO and ABC will integrate to FGWC (FGWC-PSO and FGWC-ABC) by first selecting the inertia weight effectively and efficiently in the PSO and make modifications to the formula on ABC that are expected to improve performance of FGWC. Furthermore, these methods were applied to clustering regency/municipality based on infant stunting factors because stunting related with people behaviour and their location (geography). Stunting is linear growth failure due to malnutrition during the critical growth of children. The performance of cluster formed will be compare with six different cluster evaluation: Partition Coefficient, Classification Entropy, Partition Index, Separation Index, Xie and Beny Index, dan IFV Index.

Item Type: Thesis (Masters)
Uncontrolled Keywords: ABC; FCM; FGWC; FGWC-ABC; FGWC-PSO; PSO; Fuzzy Geographically Weighted Clustering; Fuzzy C-Mean; Artificial Bee Colony; Particle Swarm Optimization; Partition Coefficient; Classification Entropy; Partition Index; Separation Index; Xie and Beny Indeks; IFV Index
Subjects: H Social Sciences > HA Statistics
Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis
Depositing User: BAMBANG SULISTYO HADI
Date Deposited: 21 Mar 2017 01:19
Last Modified: 21 Mar 2017 08:33
URI: http://repository.its.ac.id/id/eprint/2721

Actions (login required)

View Item View Item