Showing 1–1 of 1 results for author: Ushakov, R

Consistent Estimation of Mixed Memberships with Successive Projections

Authors: Maxim Panov, Konstantin Slavnov, Roman Ushakov

Abstract: This paper considers the parameter estimation problem in Mixed Membership Stochastic Block Model (MMSB), which is a quite general instance of random graph model allowing for overlap** community structure. We present the new algorithm successive projection overlap** clustering (SPOC) which combines the ideas of spectral clustering and geometric approach for separable non-negative matrix factori… ▽ More This paper considers the parameter estimation problem in Mixed Membership Stochastic Block Model (MMSB), which is a quite general instance of random graph model allowing for overlap** community structure. We present the new algorithm successive projection overlap** clustering (SPOC) which combines the ideas of spectral clustering and geometric approach for separable non-negative matrix factorization. The proposed algorithm is provably consistent under MMSB with general conditions on the parameters of the model. SPOC is also shown to perform well experimentally in comparison to other algorithms. △ Less

Submitted 14 October, 2017; v1 submitted 5 July, 2017; originally announced July 2017.

Journal ref: Complex Networks & Their Applications VI. COMPLEX NETWORKS 2017. Studies in Computational Intelligence, vol 689