Modified affinity propagation

2nd Nommensen International Conference on Technology and Engineering

Gunawan, Heru^1 ; Suwilo, Saib^2 ; Situmorang, Zakarias^3

Departement of Computer Science, Universitas Sumatera Utara, Medan, Indonesia^1

Department of Mathematics, Universitas Sumatera Utara, Medan, Indonesia^2

Universitas Katolik Santo Thomas, Medan, Indonesia^3

关键词: Affinity propagation;    Data distribution;    Number of clusters;    Prior knowledge;    Quality clustering;    Quality of clustering;    Silhouette indices;    Similarity matrix;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/420/1/012121/pdf
DOI  :  10.1088/1757-899X/420/1/012121

来源: IOP

【 摘 要 】

Affinity Propagation (AP) is exemplar-based clustering algorithm, this algorithmdoes not require prior knowledge of the number of clusters. The quality of clustering results is highly dependent on the "preference" value. Standard AP algorithm take "preference" value based on median or minimum value of similarity matrix, then the value is shared to all "preference" value on similarity matrix. This method does not give the best solution, because the value not represent the overall data structure. The Modified AP (M-AP) is proposed to resolve this problem. M-AP algorithmtake "preference" value based on data distribution on each row from similarity matrix. Experimental result show that M-AP cat outperform AP in quality clustering result based on Silhouette Index score.