ANALISIS KLASTER PERSEBARAN ANAK DISABILITAS DI INDONESIA DENGAN SELF ORGANIZING MAPS

RAKA ADI NUGROHO, NIM. 14650015 (2018) ANALISIS KLASTER PERSEBARAN ANAK DISABILITAS DI INDONESIA DENGAN SELF ORGANIZING MAPS. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (ANALISIS KLASTER PERSEBARAN ANAK DISABILITAS DI INDONESIA DENGAN SELF ORGANIZING MAPS) 14650015_BAB-I_V_DAFTAR-PUSTAKA.pdf Download (3MB) | Preview
	Text (ANALISIS KLASTER PERSEBARAN ANAK DISABILITAS DI INDONESIA DENGAN SELF ORGANIZING MAPS) 14650015_BAB-II_IV.pdf Restricted to Registered users only Download (2MB)

Abstract

Disability is an issue that resides in all parts of the world. Relates closely to the limitations and inability of an individual to exercise his or her responsibilities and gain his rights. There are 2.45% of the disability population or about 6 million people in Indonesia of the total population in Indonesia are persons with disabilities. Interesting facts, diseases and accidents cause disability by 76% of total persons with disabilities. Thus it is necessary to cluster the distribution of disability in order to facilitate stakeholders in designing targeted programs related to disability. There are many methods for categorizing large data, or often referred to as clustering techniques. Self Organizing Maps (SOM) is a method in the concept of artificial neural networks as a form of unsupervised learning in which learning does not require target or label output first. SOM itself is part of the Automatic Map on the Clustering concept, which is dedicated to the processing of highdimensional data. The SOM model will be used for the formation of disability dispersal clusters in Indonesia, which have certain characteristics based on the types of disabilities.. In this research, the experiment scenario was formed for 2-9 clusters, with initial value of 0.6 and the number of epoch 10 and 100, with the use of 21,670 data. The experiment resulted in 5 clusters as the best total cluster, with the distribution of 3265 regions in the 1st cluster, 385 regions in the 2nd cluster, 756 areas in the 3rd cluster, 17185 area in the 4th cluster, and 79 cluster-5 . Where the area belonging to the 3rd cluster is an area with red disability value, and the area belonging to the 4th cluster is a minimally disability area. This value is calculated from the davies bouldin index which states 5 clusters have the best intra-cluster proximity value that is 1.1177502242256903 at iterations 10 and 100.

Share this knowledge with your friends :

Actions (login required)

View Item