SISTEM MULTIAGEN UNTUK PENGKLASTERAN PENDAKI MENGGUNAKAN K-MEANS
| Main Authors: | , Maya Cendana, , Dr. Azhari S.N., M.T |
|---|---|
| Format: | Thesis NonPeerReviewed |
| Terbitan: |
[Yogyakarta] : Universitas Gadjah Mada
, 2014
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/130715/ http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=71142 |
Daftar Isi:
- The beginner climbers should do mountain climbing as a group, but the current classification system is not capable of automatically clustering hikers. Therefore, an application program that is able to automatically clustering climbers is needed, especially for the solo climber who does not have the climbing community. Clustering will be done with K-means clustering algorithm based on intelligent agents. The agents involved are the user agent, the database agent, clustering agent, and validation agent. The agents will collaborate to determine the best cluster for the climbers to the K-means clustering can be done once / multithread. Clustering process will go through two phases: the auction and validation phase. The agents will be built on top of the JADE platform with communication language FIPA ACL. Evaluation of the 10, 100 and 200 climbers with the data to a particular cluster number to calculate the value of cohesion / density in a cluster and inter-cluster separation distances. Metric measurement used is WGAD and BGAD.