Two density-based k-means initialization algorithms for non-metric data clustering

被引:20
作者
Bianchi, Filippo Maria [1 ]
Livi, Lorenzo [1 ]
Rizzi, Antonello [1 ]
机构
[1] SAPIENZA Univ Rome, Dept Informat Engn Elect & Telecommun, Via Eudossiana 18, I-00184 Rome, Italy
关键词
Clustering; Prototype selection; k-means initialization; Dissimilarity measures; Non-metric domains; GRAPH; KERNEL;
D O I
10.1007/s10044-014-0440-4
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this paper, we propose a density-based clusters' representatives selection algorithm that identifies the most central patterns from the dense regions in the dataset. The method, which has been implemented using two different strategies, is applicable to input spaces with no trivial geometry. Our approach exploits a probability density function built through the Parzen estimator, which relies on a (not necessarily metric) dissimilarity measure. Being a representatives extractor a general-purpose algorithm, our method is obviously applicable in different contexts. However, to test the proposed procedure, we specifically consider the problem of initializing the k-means algorithm. We face problems defined on standard real-valued vectors, labeled graphs, and finally sequences of real-valued vectors and sequences of characters. The obtained results demonstrate the effectiveness of the proposed representative selection method with respect to other state-of-the-art solutions.
引用
收藏
页码:745 / 763
页数:19
相关论文
共 43 条
[1]   Graph matching and clustering using kernel attributes [J].
Angel Lozano, Miguel ;
Escolano, Francisco .
NEUROCOMPUTING, 2013, 113 :177-194
[2]  
[Anonymous], 2013, PATTERN RECOGN, DOI DOI 10.1007/978-3-642-36530-0_7
[3]  
[Anonymous], 2007, P 18 ANN ACM SIAM S
[4]  
[Anonymous], P 2013 INT JOINT C N
[5]  
Bache K., 2013, UCI Machine Learning Repository
[6]  
Bardaji I, 2010, LECT NOTES COMPUT SC, V6218, P149, DOI 10.1007/978-3-642-14980-1_14
[7]   A Granular Computing approach to the design of optimized graph classification systems [J].
Bianchi, Filippo Maria ;
Livi, Lorenzo ;
Rizzi, Antonello ;
Sadeghian, Alireza .
SOFT COMPUTING, 2014, 18 (02) :393-412
[8]   A Game-Theoretic Approach to Hypergraph Clustering [J].
Bulo, Samuel Rota ;
Pelillo, Marcello .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2013, 35 (06) :1312-1327
[9]   Clustering by compression [J].
Cilibrasi, R ;
Vitányi, PMB .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2005, 51 (04) :1523-1545
[10]  
Del Vescovo Guido, 2014, International Journal of Computer Theory and Engineering, V6, P9, DOI 10.7763/IJCTE.2014.V6.827