Analisis Dan Penerapan Algoritma K-Means Dalam Strategi Promosi Kampus Akademi Maritim Suaka Bahari

Abstract

The process of accepting new cadet candidates at the Maritime Academy of Marine Sanctuary every year, produces a lot of data in the form of profiles of prospective cadets. The activity caused a large accumulation of data, it became difficult to identify prospective cadets. This research discusses the application of data mining to generate profiles that have similar attributes. One of the data mining techniques used to identify a group of objects that have the same characteristics is Cluster Analysis. The data clustering method is divided into one or more clusters that have the same characteristics called K-means. The method that the author uses is knowledge discovery in databases (KDD) consisting of Data, Data Cleaning, Data transformation, Data mining, Pattern evolution, knowledge. Implementation of K-means Clustering process using Rapid Miner. Attributes used by NIT, Level, Name, Student Status, Type of Registration, Gender, Place of Birth, Date of Birth, Religion, School Origin, School Origin Department, Religion, GPA, Subdistrict, District/ City, Province. Returns the number of clusters 30 (k=30). From the research results based on davies bouldin test on K-means algorithm resulted in the closest value of 0 is k = 29 with Davies bouldin: 0.070, with the most cluster member distribution in cluster 16 containing cluster members 115 items.