Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2023
- Goals
- Record
- a6f8d7d3-b26b-4d30-a07f-283664aa2c2d
- Published
- 2026-10-30
Universitas Dian Nuswantoro, through researchers Wibowo, Nanang Lestio; Soeleman, Moch Arief; and Fanani, Ahmad Zainul, conducted research on the modification of the Antlion Optimizer algorithm for determining the initial centroid in the K-means algorithm. The activity addressed clustering, which is described as the grouping of data in data mining processing. K-means is identified as a popular clustering algorithm because it is easy to use and fast in clustering data. However, the method determines the initial centroid randomly as a reference for processing, and careless centroid selection can result in poor clustering processes and local optima.
The researchers used the modified Antlion Optimizer method to improve the initial centroid determination in K-means and to provide an alternative for obtaining better clustering results. The proposed method was evaluated against the usual K-means and K-means++ methods using the Sum of Intra-Cluster distance evaluation. The evaluation used the UCI datasets named iris, wine, glass, ecoli, and cancer. The best SICD value for each method and dataset was measured and used to produce a ranking. The proposed method achieved the first rank on the iris, wine, and cancer datasets. On the ecoli and glass datasets, the proposed method and the K-means++ method both achieved the first rank. Based on the average ranking value, the proposed method was ranked first. The research therefore presents the modified Antlion Optimizer as a method that can improve clustering results and as an alternative for determining the initial cluster center in K-means. This UDINUS research addresses the stated goal of improving the determination of initial centroids in clustering.