Maxim Panov

Research Scientist

Joint affiliation with a group of Prof. Maxim Fedorov

Phone: +7 (495) 280 14 81 ext. 3504.
Office: 335.

Skolkovo Institute of Science and Technology, Skolkovo Innovation Center, Building 4, Moscow, 143026, Russian Federation.

Research Interests

  • Algorithms and statistical analysis for random graphs
  • Bayesian methods in machine learning and statistics
  • Nonparametric and semiparametric statistical inference


  • Applied Statistics (Fall 2011-2017 at Phystech, Fall 2016 at HSE)



  • Maria Burkina
  • Dmitry Ermilov
  • Alfredo de la Fuente
  • Marina Gomtsyan
  • Anastasia Koloskova
  • Evgeny Marshakov
  • Nikita Mokrov
  • Mikhail Pautov
  • Stanislav Tsepa
  • Roman Ushakov



  • Kirill Kuznetsov (MSc, HSE 2017)
  • Konstantin Slavnov (MSc, HSE 2017)
  • Anton Votinov (MSc, HSE 2017)
  • Igor Silin (BSc, MIPT 2016)

Current projects

  1. Provable overlapping community detection
  2. Sparse inductive matrix completion
  3. Graph nodes embeddings

For prospective students

Here I summarized ideas for several research projects which can be conducted under my supervision:

  1. Model selection in overlapping community detection
  2. Semi-supervised nodes classification
  3. Local message-passing algorithms for community detection and classification


  • Maxim Panov, Konstantin Slavnov and Roman Ushakov “Consistent Estimation of Mixed Memberships with Successive Projections” Proceedings of Complex Networks 2017 (The 6th International Conference on Complex Networks and Their Applications) (arXiv:1707.01350).
  • Nikita Mokrov and Maxim Panov “Simultaneous Matrix Diagonalization for Structural Brain Networks Classification” Proceedings of Complex Networks 2017 (The 6th International Conference on Complex Networks and Their Applications) (arXiv:1710.05213).
  • Dmitry Ermilov, Maxim Panov and Yury Yanovich “Automatic Bitcoin Address Clustering”, International Conference of Machine Learning and Applications, 2017.
  • Konstantin Slavnov and Maxim Panov “Overlapping Community Detection in Weighted Graphs: Matrix Factorization Approach”, Proceedings of IIP conference, Springer, 2017.

Further information

Skoltech profile
Google Scholar