Преподаватели


Шевляков Георгий Леонидович

д.ф.-м.н.
профессор
Профессор кафедры «Прикладная математика»

Georgy L. SHEVLYAKOV

PROFESSOR OF MATHEMATICS

Date of birth: 20.08.1949

Place of birth: Leningrad, USSR

WORK ADDRESS:

Department of Applied Mathematics, St. Petersburg State Polytechnic University,

Polytechnicheskaya ul., 29, St. Petersburg, 195251, Russia

Fax: +7-812-356 36 64 Phone: +7-812-552 67 50

E-mail: Этот e-mail адрес защищен от спам-ботов, для его просмотра у Вас должен быть включен Javascript

HOME ADDRESS:

Nalichnaya ul., 36-5-399, St. Petersburg, 199226, Russia

Phone:  +7-812-356 36 64

EDUCATION

  • St. Petersburg State Polytechnic University, St. Petersburg, Russia

    Doctor of Sciences in Physics and Mathematics, December 1991

    Dr. Sc. Thesis: Methods and Algorithms of Data Analysis under the Conditions of Uncertainty

  • Leningrad Polytechnic Institute, Leningrad, USSR

    Ph.D. in Statistics, June 1976

    Ph.D. Thesis: Robust Estimation and Detection of Signals by the Least Absolute Values Method

  • Leningrad Polytechnic Institute, Leningrad, USSR

    M. Sc. in Cybernetics, March 1973

    Diploma summa cum laude

    Master’s Thesis: On Applications of Order Statistics to Estimation of Signals and Rejection of Outliers

PROFESSIONAL EXPERIENCE

  • Department of Applied Mathematics, St. Petersburg State Polytechnic University, St. Petersburg, Russia

    March 2009 - present, Professor

  • Institute for Problems in Mechanical Engineering, Russian Academy of Sciences, St. Petersburg, Russia

    March 2009 - present, Principal Researcher (part time)

  • Department of Information and Communications, Gwangju Institute of Science and Technology (GIST), South Korea

    September 2003- February 2009, Visiting Professor

  • Department of Mathematics, St. Petersburg State Polytechnic University, St. Petersburg, Russia

    March 1993 - December 2006, Professor

    October 1986 - February 1993, Associate Professor

  • Institute for Problems in Mechanical Engineering, Russian Academy of Sciences, St. Petersburg, Russia

    October 1994 - August 2003, Principal Researcher (part time)

  • Department of Mechanics and Control Processes, Leningrad Polytechnic Institute, Leningrad, USSR

    October 1980 - September 1986, Senior Researcher

    May 1979- September 1980, Researcher

  • Department of Biometrics, Vavilov Research Institute for Plant Breeding, Leningrad, USSR

    October 1976 - April 1979, Junior Researcher

SUMMARY OF TEACHING EXPERIENCE

The courses delivered: Analysis (Calculus), Theoretical Mechanics, Probability Theory, Mathematical Statistics, Stochastic Processes, Queueing Theory, System Theory, Information and Coding Theory, Robust Statistics, Applied Statistics and Mathematical Methods of Data Analysis, Robust Statistical Methods in Signal Processing, Methods of Optimization and their Applications.

SUMMARY OF RESEARCH ACTIVITIES

My interests are mainly concerned with the theory and applications of robust statistical methods and with mathematical methods and technologies of data analysis. The researches aim both at theoretical aspects and real-life needs of data and signal processing.

Current Activities and Future Research Plans

These activities and plans are mainly concerned with robust statistics and data analysis technologies including the extension of robust methods on the problems of multivariate statistical and time series analysis, the study of data compression procedures, normalizing data transformations, as well as some problems in queueing theory and general statistics with their applications to data and signal processing in electrical engineering:

  • Robust multivariate statistics: robust correlation, detection of multivariate outliers in the data, robust classification.
  • Robust time-series analysis: robust detection, robust spectra estimation.
  • Reliability theory: robust estimation of scale in reliability models, application of robust methods in reliability.
  • Semantic data analysis: taxonomy evaluation, datasets evaluation, multi-label hierarchical classification.
  • Nonparametric statistics: estimation of quantile functions by the Bernstein polynomials and their generalizations.
  • Data analysis: bivariate box-plots techniques, methods for normalizing the data, and probability-free methods.
  • Queueing theory: analysis of priority queueing service discipline with applications to ATM networks and Internet.
  • General statistics: parametric estimation in finite and discrete distribution models.
  • Applications: signal processing (robust estimation and detection of signals) in electrical engineering, reliability models in mechanical engineering, data processing in cardiology and ophthalmology.

PUBLICATIONS

BOOKS:

Georgy L. Shevlyakov and Nikita O. Vilchevski, Robustness in Data Analysis: criteria and methods, VSP, Utrecht, 2002, xiv+310 pages.

Probability Topics in Mathematics. Publishing House "Ivan Fedorov", St.Petersburg, 2001, 588 p. Part 4: Mathematical Statistics, pp. 399-588. Coauthor: Yuri D. Maksimov (in Russian).

PAPERS More than 70 published papers in peer reviewed journals.

Download selected publications (4,6 Mb)

Professional Memberships

Bernoulli Society
IEEE Society


Лекционные курсы, семинары и поддержка НИРС


Руководство дипломными и выпускными работами

NСтудентГруппаКвалификация / СтепеньГодТема
1.Картелев Глеб Игоревич43601/3 Бакалавр 2016 Радикальная оценка коэффициента корелляции двумерного нормального закона распределения 
2.Шеина Екатерина Сергеевна63601/2 Магистр 2016 Робастное оценивание параметра положения распределений: некоторые новые оценки 
3.Золотых Анастасия Андреевна43601/3 Бакалавр 2015 Корреляционный критерий выбора порядка регрессионной модели 
4.Советников Максим Олегович63601/3 Магистр 2015 Робастные оценки параметра масштаба и коэффициента корреляции нормального распределения 
5.Василевский Никита Вадимович43601/3 Бакалавр 2014 Оценивание коэффициента корреляции на основе взаимной информации 
6.Лаврентьева Галина Михайловна6057/12 Магистр 2013 Методы отбраковки многомерных данных, основанные на использовании робастных двумерных боксплотов, и их приложения в многомерном статистическом анализе. 
7.Прыткова Алена Валентиновна6057/12 Магистр 2013 Сравнительный анализ эффективности применения статистических и нейронных алгоритмов для решения задачи разладки во временных рядах 
8.Девятериков Евгений Александрович53601/3 Бакалавр 2012 Применение методов кластерного анализа для прогнозирования финансовых рынков 
9.Евграфов Алексей Владиславович63601/3 Специалист 2012 Робастное оценивание спектральной плотности процесса авторегрессии на основе модифицированных уравнений Юла-Уолкера 
10.Любомищенко Николай Сергеевич63601/2 Магистр 2012 Алгоритмы робастного оценивания спектральных характеристик сигналов и их использование  
11.Матяс Игорь Васильевич43601/3 Бакалавр 2012 Сравнение алгоритмов приведения псевдокорреляционных матриц к положительно полуопределенному виду  
12.Рубцов Александр Олегович43601/3 Бакалавр 2012 Робастные алгоритмы оценивания спектра сигналов на основе формулы Винера-Хинчина  
13.Чихиркин Сергей Сергеевич43601/1 Бакалавр 2012 Робастные алгоритмы оценивания спектральной плотности процесса авторегрессии с использованием различных моделей засорения  
14.Лаврентьева Галина Михайловна43601/2 Бакалавр 2011 Использование «быстрых» высокоэффективных робастных оценок масштаба и коэффициента корреляции при построении двумерных боксплотов для графического представления данных 
15.Маслюков Максим Андреевич63601/1 Специалист 2011 Робастное оценивание корреляционных матриц: экспериментальный анализ оценок, основанных на попарных корреляциях 
16.Любомищенко Николай Сергеевич43601/2 Бакалавр 2010 Сравнение оценок таксономий методом Монте-Карло 
17.Шолуха Павел Викторович63601/2 Магистр 2010 Разработка информационной системы проведения спортивных соревнований и исследование результатов судейства на основе экспертных оценок 

Основные публикации

 
Монографии
NАвторыНазваниеИзданиеГод
1.Georgy Shevlyakov, Hannu Oja Robust Correlation: Theory and Applications.John Wiley, 352 p. 2016 
 
Статьи
NАвторыНазваниеИзданиеГод
1.Song I.Y., Shevlyakov G., Shin V. Estimation of nonlinear functions of state vector for linear systems with time-delays and uncertainties Mathematical Problems in Engineering, Hindawi Publishing Corporation, Vol. 2015, Article ID 217253, 10 pages, http://dx.doi.org/10.1155/2015/217253 2015 
2.Shevlyakov G.L., Smirnov, P.O.  M-estimators of the correlation coefficient for bivariate independent component distributionsModern Nonparametric, Robust and Multivariate Methods. Springer Lecture Notes in Mathematics (a special issue devoted to the Festschrift of Hannu Oja), Klaus Nordhausen and Sara Taskinen (eds), October 2015, pp.127-142 2015 
3.Shevlyakov G.L., Lyubomishchenko N.S., Smirnov P.O. A few remarks on robust estimation of power spectraAustrian Journal of Statistics 44, 112—121 2014 
4.Smirnov P.O., Shevlyakov G.L. Fast highly efficient and robust one-step M-estimators of scale based on Qn-estimatorComputational Statistics and Data Analysis 78, 153—158 2014 
5.G. L. Shevlyakov, V.I. Shin, S. Lee, and Kim K. Asymptotically stable detection of a weak signalInt. Journal of Adaptive Control and Signal Processing, DOI: 10.1002/acs.2405 2013 
6.Андреа К., Смирнов П.О., Шевляков Г.Л. Двумерный боксплот на основе высокоэффективных робастных оценок масштаба и корреляцииВестник Томского государственного университета, Т.1, № 22, С. 25-31. 2013 
7.Андреа К., Шевляков Г.Л. Обнаружение выбросов с помощью боксплотов, основанных на новых высокоэффективных робастных оценках масштабаНаучно-Технические Ведомости Санкт-Петербургского политехнического государственного университета. Информатика. Телекомуникации. Управление., Т.5, № 181, С. 39-45. 2013 
8.G. L. Shevlyakov, P. O. Smirnov, V. I. Shin, and Kim K. Asymptotically minimax bias estimation of the correlation coefficient for bivariate independent component distributionsJournal of Multivariate Analysis, Vol. 111, pp. 59–65 2012 
9.Jintae Park, Georgy Shevlyakov, Kiseon Kim Distributed Detection and Fusion of Weak Signals in Fading Channels with Non-Gaussian NoisesIEEE Communications Letters Vol. 16, No. 2, pp. 220-223 2012 
10.Georgy Shevlyakov et al Robust Distributed Detection with Total Power Constraint in Large Wireless Sensor NetworksIEEE Transactions on Wireless Communications, vol. 10, pp. 5, 2011.  2011 
11.Georgy Shevlyakov, Pavel Smirnov Robust estimation of the correlation coefficient: An attempt of surveyAustrian Journal of Statistics, 40(1&2):147-156, 2011 2011 
12.N.-V. Nguyen, G. Shevlyakov, V. Shin Fusion of correlated local estimates under non-Gaussian channel noiseInternational Journal of Robotics and Automation, vol. 25, no. 2, pp. 3365-3385, 2010 2010 
13.G. Shevlyakov, J.W. Lee, K.M. Lee, Vladimir Shin, K. Kim  Robust detection of a weak signal with redescending M-estimators: a comparative studyInternational Journal of Adaptive Control and Signal Processing, vol. 24, Issue 1, pp. 33-40, 2010. 2010 
14.Nga-Viet Nguyen, Georgy Shevlyakov and Vladimir Shin Robust state estimation fusion in power systemInternational Journal of Innovative Computing, Information and Control ICIC International c_ 2010 ISSN 1349-4198, vol. 6, no. 12, December 2010, pp. 5771–5784. 2010 
15.Nga-Viet Nguyen, Georgy Shevlyakov, Vladimir Shin Fusion of Correlated Local Estimates Under Non-Gaussian Channel NoiseInternational Journal of Robotics and Automation. 2009 2009 
16.G. Shevlyakov, J.W. Lee, K.M. Lee, Vladimir Shin, K. Kim Robust detection of a weak signal with redescending M-estimators: a comparative studyInternational Journal of Adaptive Control and Signal Processing. 2009 2009 
17.Nga-Viet Nguyen, Georgy Shevlyakov, Vladimir Shin Robust Fusion with Non-Gaussian Channel NoiseIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, vol. E92-A, No.5, pp.1293-1300. 2009 2009 
 
Доклады, тезисы и материалы конференций
NАвторыНазваниеИзданиеГод
1.Shevlyakov G.  Robust asymptotically stable detection of a weak signal International Conference on Robust Statistics 2015(ICORS 2015).Indian Statistical Institute, Kolkata, India. Book of Abstracts, January 12-16, 2015, pp.48-49 2015 
2.Shevlyakov G.  Robust estimation of a correlation coefficient: approaches and methodsIndo-Russian Joint Conference in Statistics and Probability, Lecture Abstracts.Indian Statistical Institute, Delhi, India, January 15-18, 2015 2015 
3.Andrea A., Shevlyakov G.L., Vassilieva N.S., Ulanov A.N. A new measure of outlier detection performance evaluationProc. of the International Conference on Machine Learning and Data Mining (MLDM 2014). Vol.1, pp. 21-28, St. Petersburg, July 16-21 2014 
4.Shevlyakov G.L., Smirnov P.O. Robust M-estimates of the correlation coefficient in bivariate independent component distribution modelsAbstracts of the International Conference on Robust Statistics (ICORS 2014), p. 31, Halle, Germany, August 11-15, 2014. 2014 
5.Andrea K., Shevlyakov G.L., Smirnov P.O. Detection of outliers with boxplotsProceedings of the 10th International Conference on Computer Data Analysis and Modeling, Vol.1, pp. 141-144. 2013 
6.Lavrentyeva G., Andrea K., Shevlyakov G. Optimizing outlier detection with the harmonic mean performance criterion.Abstracts of the International Conference on Robust Statistics 2013 (ICORS 2013), p. 29 2013 
7.Lyubomishchenko N., Andrea K., Prytkova A., Shevlyakov G. Real-life data anomaly detection with robust sliding spectral and boxplot techniquesAbstracts of the International Conference on Robust Statistics 2013 (ICORS 2013), pp. 30-31 2013 
8.Shevlyakov G., Andrea K., Choudur L., Smirnov P., Ulanov A., Vassilieva N. Robust versions of the Tukey boxplot with their application to detection of outliers2013 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2013), pp.6506 -6510 2013 
9.Shevlyakov G.L., Lyubomishchenko N.S., Smirnov P.O. Some remarks on robust estimation of power spectra. - (Invited talk).Proceedings of the 10th International Conference on Computer Data Analysis and Modeling, Vol.1, pp. 97-104.,  2013 
10.Georgy Shevlyakov Robust Correlation (invited talk)International Conference on Robust Statistics 2012 (ICORS 2012), August 06 -10, 2012, Burlington, Vermont, USA 2012 
11.Georgy Shevlyakov et al Enhancing Accuracy of Multilabel Classification by Extracting Hierarchies22nd International Workshop on Database & Expert Systems, DEXA 2011, Toulouse, France, August 2011, pp. 203-207 2011 
12.G.Shevlyakov, K.Andrea, G.Lavrentyeva, P.Smirnov Fast low-complexity bivariate boxplots based on highly efficient and robust estimates of dispersion & correlationInternational Conference on Robust Statistics (ICORS2011) Valladolid, 2011 2011 
13.G. Shevlyakov Lattice analogues of Fisher information & least informative distributions International Conference on Analytical Methods in Statistics (AMISTAMISTAT 2011), October 28-30, Prague, Czech Republic, p. 10.  2011 
14.Georgy Shevlyakov et al Similarity-Based Attribute Selection with n-Grams7th International Conference, (MLDM 2011), New-York, USA, August/September, pp. 72-76. 2011 
15.P.O. Smirnov and G.L. Shevlyakov  Highly efficient robust estimators of a correlation coefficient for bivariate independent component distributionsIn Proceedings of the International Conference on Robust Statistics (ICORS2010) Prague 2010 
16.Alexander Ulanov, Georgy Shevlyakov, Nikolay Lyubomishchenko, Pankaj Mehra, Vladimir Polutin Monte Carlo Study of Taxonomy Evaluation2010 Workshops on Database and Expert Systems Applications (DEXA 2010), pp. 164-168 2010 
17.P.O. Smirnov and G.L. Shevlyakov  On approximation of the Qn-estimate of scale by fast M-estimatesIn Proceedings of the International Conference on Robust Statistics (ICORS2010) Prague 2010 
18.N.-V. Nguyen, G. Shevlyakov, and V. Shin Power system state estimation with fusion method2010 International Conference on Electrical and Energy Systems, vol. 5, pp. 71-76 2010 
19.N.-V. Nguyen, G. Shevlyakov, and V. Shin Robust algorithm in distributed estimation fusion with correlation of local estimates2nd International Conference on Computer and Automation Engineering, vol. 5, pp. 428-431.  2010 
20.P.O. Smirnov and G.L. Shevlyakov  Robust estimation of a correlation coefficient: an attempt of surveyIn Proceedings of the 9th International Conference on Computer Data Analysis and Modeling: Complex Stochastic Data and Systems (CDAM2010), Minsk 2010 
21.G. Shevlyakov, M. Shevlyakova and Vladimir Shin A Least Favorable Multivariate Generalized Gaussian Pattern Distribution with its Application to Robust ClassificationProc. Machine Learning and Data Mining in Pattern Recognition 6th Inter. Conf., MLDM’09, Leipzig, Germany, pp. 74-78, 23~25 July, 2009 2009 
22.Nga-Viet Nguyen, Vladimir Shin,Georgy Shevlyakov Robust Algorithm in Distributed Estimation Fusion with Correlation of Local EstimatesProc. Intern. Conf. Signal Acquisition and Processing (ICSAP-2010) , 9~10 Feb. 2010, Bangalore, India 2009 
23.G. Shevlyakov, K. Lee, V. Shin, and K. Kim Robust Minimum Distance Neyman-Pearson Detection of a Weak Signal in Non-Gaussian NoiseProc. 17th European Signal Processing Conf. (EUSIPCO’09), 24~28 Aug. 2009, Glasgow, Scotland, pp. 1012~1016 2009 
24.G.L. Shevlyakov, V.I. Shin, and K.S. Kim The Most B- and V-robust Estimators of a Correlation CoefficientProc. Inter. Conf. Robust Statistics (ICORS’09), 14~19 June 2009, Parma, Italy, pp. 140-141 (ISBN-13: 978-88-903330-0-2) 2009