Idioma:

Parallel Spectral Clustering in Distributed Systems

Chen, Wen-Yen ; Song, Yangqiu ; Bai, Hongjie ; Lin, Chih-Jen ; Chang, Edward Y.

IEEE transactions on pattern analysis and machine intelligence, 2011-03, Vol.33 (3), p.568-586 [Periódico revisado por pares]

Los Alamitos, CA: IEEE

Texto completo disponível

Citações Citado por

Enviar para

Título:
Parallel Spectral Clustering in Distributed Systems
Autor: Chen, Wen-Yen ; Song, Yangqiu ; Bai, Hongjie ; Lin, Chih-Jen ; Chang, Edward Y.
Assuntos: Algorithms ; Applied sciences ; Approximation ; Artificial Intelligence ; Cluster Analysis ; Clustering ; Clustering algorithms ; Computation ; Computer Communication Networks - instrumentation ; Computer science ; Computer science; control theory; systems ; Computer Simulation ; Computer systems and distributed systems. User interface ; Concurrent computing ; Data processing. List processing. Character string processing ; Distributed computing ; Exact sciences and technology ; Information retrieval. Graph ; Laplace equations ; Memory organisation. Data processing ; Models, Statistical ; Nearest neighbor searches ; nearest neighbors ; normalized cuts ; Nyström approximation ; Parallel algorithms ; Parallel processing ; Parallel spectral clustering ; Pattern Recognition, Automated - methods ; Reproducibility of Results ; Scalability ; Similarity ; Software ; Sparse matrices ; Spectra ; Strategy ; Systems Integration ; Theoretical computing ; USA Councils
É parte de: IEEE transactions on pattern analysis and machine intelligence, 2011-03, Vol.33 (3), p.568-586
Notas: ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ObjectType-Article-1
ObjectType-Feature-2
Descrição: Spectral clustering algorithms have been shown to be more effective in finding clusters than some traditional algorithms, such as k-means. However, spectral clustering suffers from a scalability problem in both memory use and computational time when the size of a data set is large. To perform clustering on large data sets, we investigate two representative ways of approximating the dense similarity matrix. We compare one approach by sparsifying the matrix with another by the Nyström method. We then pick the strategy of sparsifying the matrix via retaining nearest neighbors and investigate its parallelization. We parallelize both memory use and computation on distributed computers. Through an empirical study on a document data set of 193,844 instances and a photo data set of 2,121,863, we show that our parallel algorithm can effectively handle large problems.
Editor: Los Alamitos, CA: IEEE
Idioma: Inglês

Links

Voltar para lista de resultados

Realização: Logos de Redes Sociais:

Parallel Spectral Clustering in Distributed Systems

Chen, Wen-Yen ; Song, Yangqiu ; Bai, Hongjie ; Lin, Chih-Jen ; Chang, Edward Y.

Los Alamitos, CA: IEEE

Buscando em bases de dados remotas. Favor aguardar.