People

Dr Igor Rodionov

Senior Lecturer
School of Mathematics, Statistics and Actuarial Science (SMSAS)
Dr Igor Rodionov

Profile

Biography

I'm a Senior Lecturer at the School of Mathematics, Statistics and Actuarial Science of the University of Essex. Previously I held a research position at Ecole Polytechnique Federale de Lausanne (Switzerland), also I was a full professor at MIPT (Russia) and leading researcher at the Russian Academy of Sciences. My research is on extreme value theory, from theoretical foundations of this theory to its applications. More specifically, now I work on problems in the following areas: *) Survival analysis for extreme events; *) Hypotheses testing and model selection for univariate extremes; *) Theory of phantom distribution functions on various spaces; *) The block quantile method for fitting models for extremes and its applications to environmental data.

Qualifications

  • Doctor of Science (Russian analog of Habilitation) Steklov Mathematical Institute, (2021)

  • PhD Lomonosov Moscow State University, (2014)

  • MSc Lomonosov Moscow State University, (2010)

Appointments

University of Essex

  • Senior Lecturer, School of Mathematics, Statistics and Actuarial Science, University of Essex (1/8/2023 - present)

Other academic

  • Scientist, Chair of Statistics, École Polytechnique Fédérale de Lausanne (1/11/2022 - 31/7/2023)

  • Full Professor, part-time, Department of Discrete Mathematics, Moscow Institute of Physics and Technology (9/2021 - 9/2022)

  • Leading Researcher, Russian Academy of Sciences (1/2/2022 - 30/9/2022)

  • Senior Researcher, Russian Academy of Sciences (1/4/2019 - 31/1/2022)

  • Associate Professor, part-time, Department of Discrete Mathematics, Moscow Institute of Physics and Technology (9/2014 - 8/2021)

  • Senior Researcher, Department of Probability Theory, Lomonosov Moscow State University (1/3/2016 - 31/3/2019)

  • Junior Researcher, Department of Probability Theory, Lomonosov Moscow State University (1/10/2012 - 29/2/2016)

  • Assistant Professor, part-time, Department of Discrete Mathematics, Moscow Institute of Physics and Technology (1/9/2012 - 31/8/2014)

Research and professional activities

Research interests

Statistics of extremes

I'm interested in a wide range of topics in statistics of extremes from theoretical aspects to practical applications.

Key words: Block quantiles method, hypotheses testing for extremes, Environmental extremes, Extremal dependence, Extreme value index, Multivariate extremes
Open to supervise

Extreme value theory

Currently, I focus on two directions of probabilistic extreme value theory: investigation of the properties of phantom distribution functions in various spaces and their applications in statistics, and establishing extreme survival analysis which includes proving central limit theorem for censored extremes, study of the properties of extreme Kaplan-Meyer estimator etc.

Key words: CLT for extremes, Extreme Kaplan-Meyer estimation, Phantom distribution, Random censoring, Extremes of Random Fields
Open to supervise

Current research

Censored extreme value estimation

A novel and comprehensive methodology designed to tackle the challenges posed by extreme values in the context of random censorship is introduced. The main focus of this study revolves around the analysis of integrals based on the product-limit estimator of normalized top-order statistics, denoted extreme Kaplan–Meier integrals. These integrals allow for the derivation of various important asymptotic distributional properties, offering an alternative approach to conventional plug-in estimation methods. Notably, this methodology demonstrates robustness and wide applicability within the scope of max-domains of attraction. An additional noteworthy by-product is the extension of residual estimation of extremes to encompass all max-domains of attraction, which is of independent interest.

Location and scale free selection between distribution tail models

We consider the problem of distinguishing between two distribution tail models, under the assumption that tails from one model are lighter than tails from another. Except the distribution that we use for separating these two models, we do not assume that the distributions belong to any of the maximum domains of attraction. Two tests are proposed: a scale free and a scale and location free, and their asymptotic properties are established. The efficiency of the developed tests in comparison with the tests proposed in the literature is examined on the basis of a simulation study. We also apply our tests to real data on human mortality at extreme age in US.

Conferences and presentations

Estimation of extremes of long precipitation series via the block quantile method

Sixth international conference on Advances in Extreme Value Analysis and Application to Natural Hazard (EVAN), Venice, Italy, 18/7/2024

Estimation of extreme quantiles via the block quantile method with application to long precipitation series

STOR-i Extremes Workshop (STEW 2023), Lancaster University, United Kingdom, 22/9/2023

Quantile estimation for maxima of stationary sequence by the block quantile method

European Meeting of Statisticians 2023, University of Warsaw, Poland, 6/7/2023

Beyond Generalized Pareto: testing and selection of tail models

13th conference on Extreme Value Analysis, Probabilistic and Statistical Models and their Applications (EVA 2023 Milano), Bocconi University, Milan, Italy, 26/6/2023

Beyond Generalized Pareto: how to select the model for distribution tail?

German Probability and Statistics Days, Essen, 2023, University of Duisburg-Essen, Germany, 8/3/2023

Precise large deviations for m-dependent subexponential sequences

The 12th international conference in extreme value analysis, methods and its applications (EVA 2021 Edinburgh), Edinburgh, United Kingdom, 2/7/2021

On threshold selection problem for estimation of extremal index

5TH INTERNATIONAL CONFERENCE ON STOCHASTIC METHODS (ICSM-5), Moscow, Russia, 23/11/2020

Extreme value theory for dependent random variables with application to random graphs

The 8th Polish Combinatorial Conference, Poland, 16/9/2020

Discrepancy method for extremal index estimation

Extreme Value Analysis 2019 (EVA 2019 Zagreb), Zagreb, Croatia, 2/7/2019

On discrimination between two classes of distribution tails

4TH CONFERENCE OF THE INTERNATIONAL SOCIETY FOR NONPARAMETRIC STATISTICS, Fisciano, Italy, 13/6/2018

Teaching and supervision

Current teaching responsibilities

  • Stochastic Processes (MA319)

  • Capstone Project: Mathematics (MA829)

  • Capstone Project: Mathematics (MA830)

  • Capstone Project: Mathematics (MA831)

Publications

Journal articles (18)

Isaev, M., Rodionov, I., Zhang, R-R. and Zhukovskii, M., Extremal independence in discrete random systems. Annales de l’Institut Henri Poincaré - Probabilités et Statistiques

Davison, AC. and Rodionov, I., (2024). Anthony C Davison and Igor Rodionov’s contribution to the Discussion of ‘Estimating means of bounded random variables by betting’ by Waudby-Smith and Ramdas. Journal of the Royal Statistical Society Series B: Statistical Methodology. 85 (1), 35-36

Rodionov, IV. and Zhukovskii, ME., (2023). The distribution of the maximum number of common neighbors in the random graph. European Journal of Combinatorics. 107, 103602-103602

Markovich, N. and Rodionov, I., (2023). Threshold selection for extremal index estimation. Journal of Nonparametric Statistics. 36 (3), 527-546

Kantonistova, EO. and Rodionov, IV., (2022). On Procedures for Testing the Equivalence of Distribution Tails. Doklady Mathematics. 106 (3), 436-439

Rodionov, I. and Vaičiulis, M., (2022). The Estimation of Parameters for the Tapered Pareto Distribution from Incomplete Data. Lithuanian Mathematical Journal. 62 (3), 391-411

Kogut, NS. and Rodionov, IV., (2021). On Tests for Distinguishing Distribution Tails. Theory of Probability & Its Applications. 66 (3), 348-363

Piterbarg, VI. and Rodionov, IV., (2021). Certain modern developments in stochastic extreme value theory, on occasion of 110th birthday of Boris Vladimirovich Gnedenko. Reliability: Theory and Applications. 16 (4), 16-25

Kobelkov, SG., Piterbarg, VI. and Rodionov, IV., (2021). Correction to: On maximum of Gaussian random fields having unique maximum point of its variance. Extremes. 24 (1), 85-90

Mikosch, T. and Rodionov, I., (2021). Precise large deviations for dependent subexponential variables. Bernoulli. 27 (2), 1319-1347

Jakubowski, A., Rodionov, I. and Soja-Kukieła, N., (2021). Directional phantom distribution functions for stationary random fields. Bernoulli. 27 (2), 1028-1056

Markovich, NM. and Rodionov, IV., (2020). Maxima and sums of non-stationary random length sequences. Extremes. 23 (3), 451-464

Rodionov, IV. and Sozontov, AN., (2020). On Confidence Estimation Based on Quantitative Similarity Coefficients. Automation and Remote Control. 81 (2), 320-332

Isaev, M., Rodionov, IV., Zhang, R-R. and Zhukovskii, ME., (2020). Extreme value theory for triangular arrays of dependent random variables. Russian Mathematical Surveys. 75 (5), 968-970

Piterbarg, VI. and Rodionov, IV., (2020). High excursions of Bessel and related random processes. Stochastic Processes and their Applications. 130 (8), 4859-4872

Rodionov, IV., (2019). Discrimination of Close Hypotheses about the Distribution Tails Using Higher Order Statistics. Theory of Probability & Its Applications. 63 (3), 364-380

Rodionov, IV., (2018). A Discrimination Test for Tails of Weibull-Type Distributions. Theory of Probability & Its Applications. 63 (2), 327-335

Rodionov, IV., (2018). On Discrimination between Classes of Distribution Tails. Problems of Information Transmission. 54 (2), 124-138

Grants and funding

2024

LMS Scheme 4 Research in Pairs - re University of Copenhagen visit

London Mathematical Society

Contact

igor.rodionov@essex.ac.uk
+44 (0) 1206 872341

Location:

3A.535, Colchester Campus

More about me