Daniil Ryabko
daniil@ryabko.net

SequeL lab of INRIA LNE, Lille, France.
currently on sabbatical at
INRIA Chile,
Santiago de Chile




Selected publications

2014

R. Ortner, O. Maillard, D. Ryabko, Selecting Near-Optimal Approximate State Representations in Reinforcement Learning. bib  |  pdf
In Proceedings of ALT, LNCS 8776, pp. 140-154, Bled, Slovenia, 2014.
R. Ortner, D. Ryabko, P. Auer and R. Munos, Regret Bounds for Restless Markov Bandits. bib  |  pdf
Theoretical Computer Science, 558:62-76, 2014.
A. Khaleghi, D. Ryabko, Asymptotically consistent estimation of the number of change points in highly dependent time series. bib  |  pdf
In Proceedings of ICML, JMLR W&CP 32(1):539-547, Beijing, China, 2014.

2013

D. Ryabko, J. Mary, A Binary-Classification-Based Metric between Time-Series Distributions and Its Use in Statistical and Learning Problems. bib  |  pdf
Journal of Machine Learning Research, vol. 14: 2837-2856, 2013.
A. Khaleghi, D. Ryabko, Nonparametric multiple change point estimation in highly dependent time series. bib  |  pdf
In Proceedings of ALT, LNCS 8139, pages 382-396, Singapore, 2013.
D. Ryabko, Unsupervised model-free representation learning. bib  |  pdf (tr version)
In Proceedings of ALT, LNCS 8139, pages 354-366, Singapore, 2013.
D. Ryabko, Time-series information and learning. bib  |  pdf
In Proceedings of ISIT, pages 1392-1395, Istanbul, Turkey, 2013.
O. Maillard, P. Nguyen, R. Ortner, D. Ryabko, Optimal Regret Bounds for Selecting the State Representation in Reinforcement Learning. bib  |  pdf
In Proceedings of ICML, JMLR W&CP 28(1):543-551, Atlanta, USA, 2013.

2012

D. Ryabko, Testing composite hypotheses about discrete ergodic processes. bib  |  pdf
Test, vol. 21(2), pp. 317-329, 2012.
D. Ryabko, Uniform hypothesis testing for finite-valued stationary processes. bib  |  pdf
Statistics vol. 48(1), pp. 121-128, 2014.
A. Khaleghi, D. Ryabko, Locating Changes in Highly Dependent Data with Unknown Number of Change Points. bib  |  pdf
In Proceedings of NIPS, pp. 3095--3103, Lake Tahoe, USA, 2012.
R. Ortner, D. Ryabko, Online Regret Bounds for Undiscounted Continuous Reinforcement Learning. bib  |  pdf
In Proceedings of NIPS, pp. 1772--1780, Lake Tahoe, USA, 2012.
A. Khaleghi, D. Ryabko, J. Mary, P. Preux, Online Clustering of Processes bib  |  pdf
In Proceedings of AISTATS 2012, La Palma, Canary Islands, Spain, JMLR W&CP 22: 601-609, 2012.

2011

D. Ryabko, On the relation between realizable and non-realizable cases of the sequence prediction problem. bib  |  pdf
Journal of Machine Learning Research, vol. 12: 2161-2180, 2011.
B. Ryabko, D. Ryabko, Constructing perfect steganographic systems, bib  |  pdf
Information and Computation, 2011, Vol. 209, No. 9, pp. 1223-1230.
O. Maillard, R. Munos, D. Ryabko, Selecting the State-Representation in Reinforcement Learning bib  |  pdf
In Proceedings of NIPS, Granada, Spain, pp. 2627-2635, 2011.
B. Ryabko, D. Ryabko, Confidence Sets in Time-Series Filtering. bib  |  pdf
In Proceedings of IEEE International Symposium on Information Theory (ISIT'11), St. Petersburg, Russia, pp. 2436-2438, 2011.

2010

D. Ryabko, On Finding Predictors for Arbitrary Families of Processes. bib  |  pdf
Journal of Machine Learning Research, vol. 11(Feb): 581-602, 2010.
D. Ryabko, B. Ryabko Nonparametric Statistical Inference for Ergodic Processes, bib  |  pdf
IEEE Transactions on Information Theory, 56(3):1430-1435, 2010.
D. Ryabko, Discrimination between B-processes is impossible, bib  |  pdf
Journal of Theoretical Probability, 23(2):565-575, 2010.
D. Ryabko. Sequence prediction in realizable and non-realizable cases, bib  |  pdf
In Proceedings of The 23rd Annual Conference on Learning Theory (COLT), Haifa, Israel, pp. 119-131, 2010.
D. Ryabko. Clustering processes, bib  |  pdf
In Proceedings of 27th International Conference on Machine Learning (ICML), Haifa, Israel, pp. 919-926, 2010

2009 and before

selected journal papers
B. Ryabko, D. Ryabko, Asymptotically Optimal Perfect Steganographic Systems. bib  |  pdf
Problems of Information Transmission, 2009, Vol. 45, No. 2, pp. 184-190.
D. Ryabko, J. Schmidhuber Using data compressors to construct order tests for homogeneity and component independence, bib  |  pdf
Applied Mathematics Letters, 22:7, 1029-1032, 2009.
D. Ryabko, M. Hutter. On the Possibility of Learning in Reactive Environments with Arbitrary Dependence. bib  |  pdf
Theoretical Computer Science, Vol. 405, pp. 274-284, 2008.
D. Ryabko, M. Hutter. Predicting Non-Stationary Processes. bib  |  tr pdf
Applied Mathematics Letters Vol. 21(5) pp 477-482, 2008.
D. Ryabko. On sample complexity for computational classification problems, bib  |  pdf
Algorithmica, 49:1 (Sept): 69-77, 2007.
D. Ryabko. Pattern Recognition for Conditionally Independent Data, bib  |  pdf
Journal of Machine Learning Research 7(Apr):645-664, 2006.
A. Gutman, D. Ryabko Nonstandard hull of a normed space in a Boolean-valued universe. bib  |  pdf
Siberian Advances in Mathematics, Vol. 12 No 2, pp 38-47, 2002.

Theses

D. Ryabko. Learnability in Problems of Sequential Inference (HDR Thesis) bib  |  pdf
Université Lille 1 - Sciences et Technologies, 2011.
D. Ryabko. On the Flexibility of Theoretical Models for Pattern Recognition (Ph.D. Thesis) bib  |  pdf
Royal Holloway, University of London. April, 2005.
D. Ryabko On Some Properties of Continuous Polyverse (Ph.D. Thesis) bib  |  pdf (in Russian)
Novosibirsk State University, Sobolev Institute of Mathematics; Novosibirsk, Russia, 2003.

Selected students

Azadeh Khaleghi, On Some Unsupervised Learning Problems for Highly Dependent Time Series, pdf
PhD 2010-2013.