Correlation clustering: divide and conquer

The correlation clustering is an NP-hard problem, hence its solving methods do not scale well. The contraction method and its improvement enable us to construct a divide and conquer algorithm, which could help us to clustering bigger sets. In this article we present the contraction method and compare the effectiveness of this new new and our old methods.

[1]  Anindya Bhattacharya,et al.  Divisive Correlation Clustering Algorithm (DCCA) for grouping of genes: detecting varying patterns in expression profiles , 2008, Bioinform..

[2]  László Aszalós,et al.  Correlation Clustering by Contraction, a More Effective Method , 2016 .

[3]  Tamás Mihálydeák,et al.  Rough Classification Based on Correlation Clustering , 2014, RSKT.

[4]  Tamás Mihálydeák,et al.  Correlation clustering by contraction , 2015, 2015 Federated Conference on Computer Science and Information Systems (FedCSIS).

[5]  Avrim Blum,et al.  Correlation Clustering , 2004, Machine Learning.

[6]  Inge Li Gørtz,et al.  Union-Find with Constant Time Deletions , 2005, ICALP.

[7]  Z. Néda,et al.  Phase transition in an optimal clusterization model , 2006 .

[8]  J. Zahn.  Approximating Symmetric Relations by Equivalence Relations , 1964 .

[9]  Zhikui Chen,et al.  A clustering approximation mechanism based on data spatial correlation in wireless sensor networks , 2010, 2010 Wireless Telecommunications Symposium (WTS).

[10]  Tamás Mihálydeák,et al.  Rough Clustering Generated by Correlation Clustering , 2013, RSFDGrC.

[11]  Thomas DuBois.  Improving Recommendation Accuracy by Clustering Social Networks with Trust , 2009 .

[12]  Sebastian Nowozin,et al.  Higher-Order Correlation Clustering for Image Segmentation , 2011, NIPS.

[13]  Jiming Liu,et al.  Community Mining from Signed Social Networks , 2007, IEEE Transactions on Knowledge and Data Engineering.