If you made any changes in Pure these will be visible here soon.

Personal profile

Research interests

My current research is in the area of machine learning, including both mathematical foundations and applications. Prediction with expert advice and topological data analysis are my main topics of interest. 

Prediction with expert advice is a paradigm for sequential forecasting that studies how one can merge predictions from different sources when reliability of these sources needs to be inferred from the predictions themselves. This paradigm is close to online bandits and related to many other machine learning techniques, from ridge regression to boosting. I am particularly interested in direct applications for prediction with expert advice algorithms.

Topological data analysis is a relatively recent but already mature area of machine learning that tries to apply notions from topology, a highly abstract branch of pure mathematics, to finding patterns in point clouds. Currently, I am working mostly on expressing topological data analysis ideas in the form of kernels inducing a reproducing kernel Hilbert space structure on data.

I am also remaining interested in the developments in both of my original research areas: algorithmic information theory / Kolmogorov complexity and constructive logical semantics.

Approach to teaching

I have taught a range of mathematical and statistical courses for both undergraduate and postgraduate students. My ideal teaching includes a rigorous but concise mathematical presentation of the main ideas illustrated by a few carefully selected examples and extensive computer simulations where possible. However, I believe that the only way to learn something is to do it, and hence I aim at providing students with a wide range of exercises, from simple modifications of lecture examples to real-world (or quasi real-world) problems that require combining and synthesising knowledge from different sources.

Scholarly biography

I studied mathematics at Moscow State University and completed a PhD in algorithmic information theory and non-classical logics in 2003. Then my research interests shifted to machine learning and I held research positions in IDSIA (Lugano, Switzerland), LIFM (Marseille, France), Royal Holloway University of London and Durham University. Finally, after a lectureship at University of Bedfordshire, I joined University of Brighton in 2015.

Supervisory Interests

I am interested in supervising postgraduate research students in sequential online forecasting, particularly prediction with expert advice.

Fingerprint Dive into the research topics where Alexey Chernov is active. These topic labels come from the works of this person. Together they form a unique fingerprint.

Algorithmic Complexity Mathematics
Prediction Error Mathematics
Learning systems Engineering & Materials Science
Computability Mathematics
Monotone Mathematics
Semilattice Mathematics
Prefix Mathematics
Kolmogorov Complexity Mathematics

Network Recent external collaboration on country level. Dive into details by clicking on the dots.

Research Output 2002 2016

  • 8 Conference contribution with ISSN or ISBN
  • 5 Article

A closer look at adaptive regret

Adamskiy, D., Koolen, W., Chernov, A. & Vovk, V., 1 Apr 2016, In : The Journal of Machine Learning Research. 17, 23, p. 1-21 21 p.

Research output: Contribution to journalArticle

Open Access

An Investigation on Online Versus Batch Learning in Predicting User Behaviour

Burlutskiy, N., Petridis, M., Fish, A., Chernov, A. & Ali, N., 5 Nov 2016, International Conference on Innovative Techniques and Applications of Artificial Intelligence. Cambridge, UK: Springer International Publishing, p. 135-149 15 p.

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBN

Open Access
Learning algorithms
Bayesian networks
Learning systems
Deep learning

Specialist experts for prediction with side information

Kalnishkan, Y., Adamskiy, D., Chernov, A. & Scarfe, T., 4 Feb 2016, Proccedings of the 2015 IEEE 15th International Conference on Data Mining Workshops. IEEE, p. 1470-1477 8 p. (IEEE International Conference on Data Mining (ICDM)).

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBN

Open Access

Reconstructing persistent graph structures from noisy images

Chernov, A. & Kurlin, V., 15 Mar 2013, In : Image-A: Applicable Mathematics in Image Engineering. 3, 5, p. 19-22 4 p.

Research output: Contribution to journalArticle

Open Access
Learning systems

Prediction with Expert Advice under Discounted Loss

Chernov, A. & Zhdanov, F., 31 Dec 2010, 21st International Conference, ALT 2010. Berlin: Springer, Vol. 6331. p. 255-269 15 p. (Lecture Notes in Computer Science).

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBN