Data Mining, Evaluation, K-means Evaluation of Supporting Work Quality Using K-Means Algorithm

Abstract

One of the factors that to improve the performance productivity of an organization or agency is Human Resources. During this time many government agencies that do not have employees with adequate competence, this is evidenced by the low productivity of employees and the difficulty of measuring employee performance in the scope of government agencies. This research discusses the application of K-means Clustering conducted at Udayana University, especially on annual performance data for contract workers. this study aims to classify cluster clusters that are determined to facilitate the evaluate quality of work of contract workers. This research uses data used as many as 1613 data. And done preprocessing get 544 data. In preprocessing data, K-means Clustering method is performed. In K-means Clustering determined the number of K as much as 5 Cluster. To determine Data Cluster used Ecludian Distance calculation. From the results of K-means Clustering Applying it takes 10 iterations. From 5 clusters conducted on 544 data there are clusters 0 as much as 38, Cluster 1 as much as 473, Cluster 2 as much as 130, Cluster 3 as much as 26 and from cluster 4 as many as 3. From the results of K-means Clustering implementation is used as a supporter of Quality Evaluation Work.