Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : Jurnal INFOTEL

Enhancement of K-Parameter Using Hybrid Stratified Sampling and Genetic Algorithm Rima Dias Ramadhani; Agus Priyanto; Muhammad Fajar Sidiq
JURNAL INFOTEL Vol 10 No 1 (2018): February 2018
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/infotel.v10i1.343

Abstract

Clustering is a technique used to classify data into clusters based on their similarities. K-means is a clustering algorithm method that classifies the objects based on their closest distance to the cluster center to the groups that have most similarities among the members. In addition, K-means is also the most widely used clustering algorithm due to its ease of implementation. However, the process of selecting the centroid on K-means still randomly. This results K-means is often trapped in local minimum conditions. Genetic algorithm is used in this research as a metaheuristic method where the algorithm can support K-means in reaching global optimum function. Besides, the stratified sampling is also used in this research, where the sampling functions by dividing the population into homogeneous areas using stratification variables. The validation value of the proposed method with iris dataset is 0.417, while the K-means is only 0.662.