Abstract | ||
---|---|---|
Data clustering is one of the most important techniques in data analysis. Although the k-means clustering method has been widely used due to its simplicity and easiness of implementation, the performance of the method depends on the initial solution, having the drawback of getting locally optimal solutions. In this paper, to solve this issue, we have proposed a data clustering method based on the firefly algorithm combined with the k-means clustering method for data clustering. In our method, the firefly algorithm first attempts to find the quasi optimal solution. Then, given the solution obtained by the firefly algorithm as an initial solution, k-means method make data clustering converge to a final solution, or final clustered data set. We demonstrate that the proposed method can be effective for data clustering using some popular benchmark data sets. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1109/ICIT.2016.7474844 | PROCEEDINGS 2016 IEEE INTERNATIONAL CONFERENCE ON INDUSTRIAL TECHNOLOGY (ICIT) |
Field | DocType | Citations |
Fuzzy clustering,Canopy clustering algorithm,Data mining,CURE data clustering algorithm,Data stream clustering,Correlation clustering,Computer science,Determining the number of clusters in a data set,Cluster analysis,Single-linkage clustering | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Kazunori Mizuno | 1 | 42 | 10.55 |
Shiho Takamatsu | 2 | 0 | 0.34 |
Toshitsugu Shimoyama | 3 | 0 | 0.34 |
Seiichi Nishihara | 4 | 71 | 14.35 |