Pengelompokan kabupaten / kota di Jawa Timur berdasarkan indikator kemiskinan menggunakan metode c-means dan fuzzy c-means clustering

Dewi, Anggraeni Rahma (2015) Pengelompokan kabupaten / kota di Jawa Timur berdasarkan indikator kemiskinan menggunakan metode c-means dan fuzzy c-means clustering. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 1311100009-Undergraduate.pdf]
Preview
Text
1311100009-Undergraduate.pdf

Download (3MB) | Preview

Abstract

Permasalahan yang sering dihadapi oleh pemerintah/negara Indone-sia adalah kemiskinan. Jika kemiskinan dapat direduksi secara drastis, ma-ka secara nasional kemiskinan akan berkurang. Jawa Timur lebih berpelu-ang dalam pengurangan jumlah angka kemiskinan. Selama ini BPS dan BAPPEDA mengelompokkan kabupaten/kota di Jawa Timur berdasarkan 4 kultur wilayah. Oleh karena itu peneliti tertarik untuk mengelompokkan ka-bupaten/kota di Jawa Timur berdasarkan indikator kemiskinan mengguna-kan metode non-hierarchical yang meliputi c-means (CM) dan fuzzy c-me-ans (FCM) clustering dengan 2 sampai 5 cluster. Hal ini dikarena adanya pendugaan kondisi kemiskinan di Jawa Timur yang belum homogen dan mengingat pengelompokan sebelumnya berjumlah 4. Pada penelitian ini di-peroleh kondisi optimum dengan kedua metode berdasarkan nilai pseudo-f statistics terbesar sebanyak 2 cluster. Metode CM clustering merupakan metode yang terbaik dalam kasus ini karena memiliki nilai icdrate dan SSW terkecil serta SSB terbesar. Melalui pengujian one-way MANOVA meng-hasilkan bahwa terjadi perbedaan karakteristik antar cluster terhadap res-pon yang dalam hal ini adalah seluruh indikator kemiskinan. Selanjutnya, pada pengujian one-way ANOVA menghasilkan bahwa terjadi perbedaan karakteristik terkait variabel lama sekolah, akses sanitasi, jenis lantai ru-mah, bahan bakar memasak, dan asset rumah. Sedangkan variabel terkait partisipasi sekolah, akses listrik, dan akses air minum karakteristik pada masing-masing cluster sama.

=========================================================================================================

Most problems have encountered by government / state of Indonesia is poverty. If poverty can be reduced drastically, then the national poverty will be reduced. East Java is more likely in a reduction in the amount of poverty. During BPS and BAPPEDA classify districts / cities in East Java by four culture regions. Therefore, researchers are interested to classify districts / cities in East Java based on poverty indicators using non-hierarchical methods that include c-means (CM) and fuzzy c-means (FCM) clustering by 2 to 5 clusters. This is caused by the estimation of poverty in East Java that has not been homogeneous and considering the previous grouping amounted to 4. In this research, the optimum conditions by both methods based on the largest pseudo-f statistics as much as 2 clusters. CM clustering method is the best method in this case because it has the smallest icdrate and SSW and largest SSB. Through the one-way MANOVA test has result in that there is difference characteristic between cluster and response which is all poverty indicators. Furthermore, the one-way ANOVA test there is difference characteristic of school period variable, sanitation access, types of floor’s house, cooking fuel, and assets of the house. While that have same characteristic in each cluster are school participation, electricity access, and drinking access.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.53 Dew p
Uncontrolled Keywords: CM clustering; FCM clustering; Icdrate; Indikator Ke-miskinan; Pseudo-f statistics
Subjects: Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: - Taufiq Rahmanu
Date Deposited: 19 Jun 2019 02:41
Last Modified: 19 Jun 2019 02:41
URI: http://repository.its.ac.id/id/eprint/63152

Actions (login required)

View Item View Item