Kuswantoro, Endik (2016) Komparasi Gabungan Algoritma Average Linkage Dan K-Means Dengan Kmeans Clustering Untuk Analisa Faktor Pengangguran Dan Angkatan Kerja - Comparison Of Algorithms Average Linkage and K-Means Clustering with K-Means Clustering to Analyze the Unemployment Factor on Work Force. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2214206706-Master thesis.pdf - Published Version Download (1MB) | Preview |
Abstract
Masalah pengangguran selalu jadi masalah bagi setiap daerah, di Jawa Timur
tingkat pengangguran terbuka menurut data BPS selalau 2 Tahun terakhir mengalami
kenaikan, hal itu disinyalir kebijakan pemerintah yang tidak tepat sasaran, dan faktor
pengangguran yang bersifat dinamis setiap waktu. Pada penelitian ini akan
menganalisa dengan mengkelompokkan faktor pengangguran dengan metode
clustering, sehingga dalam analisa faktor pengangguran diperlukan pemilihan metoda
yang tepat dan efektif. Algoritma k-means merupakan metode clustering yang mudah
diimplementasikan akan tetapi Algoritma K-Means mempunyai kekurangan pada
waktu membangkitkan nilai centroid awal yang menyebabkan hasil dari
pengelompokan terkadang baik terkadang hasilnya jelek. Pada penelitian ini penentuan
centroid awal K-Means dengan menggunakan Average Linkage clustering sebagai
salah satu alternatifnya. Dari hasil uji coba perbandingan dari dataset faktor
pengangguran nilai Silhoutte index nilai validasi Average Linkage dan K-Means
mengalami peningkatan nilai validasi 9,01 % sampai 27,4 % dari nilai K-Means murni.
========================================================================================================================
The problem of unemployment is always a problem for every region, in East
Java, the open unemployment rate according to BPS data over the last 2 years has
increased, it pointed out the government's policy which is not effective, and the factors
of unemployment that is dynamic every time, resulting in a factor analysis of
unemployment is required the selection of appropriate methods and effective. K-means
algorithm is a clustering method that is easy to implement, but the K-Means algorithm
has shortcomings at the time of initial centroid generate a value which causes the result
of the grouping is sometimes good sometimes bad result. In this study, the
determination of the initial centroid K-Means is clustering using Average Linkage as
one alternative. From the test results of a comparative dataset unemployment factor
Silhouette index value Average value validation Linkage and K-Means increases in
value validation 9.01% to 27.4% of the value of the K-Means pure
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTE 519.53 Kus k |
| Uncontrolled Keywords: | k-means, Average Linkage, Penentuan centroid awal, Faktor pengangguran, Silhoutte index, k-means, Average Linkage, initial centroid, Factor of unemployment, Silhoutte index |
| Subjects: | Q Science > QA Mathematics > QA278.55 Cluster analysis |
| Divisions: | Faculty of Industrial Technology > Electrical Engineering |
| Depositing User: | ansi aflacha |
| Date Deposited: | 17 Dec 2019 04:27 |
| Last Modified: | 17 Dec 2019 04:27 |
| URI: | http://repository.its.ac.id/id/eprint/72394 |
Actions (login required)
 |
View Item |