Homepage for Aleksey S. Polunchenko

Last Updated: June 29, 2009

Ph.D.

Present Mailing Address

University of Southern California Office: DRB 308
Department of Mathematics Tel: +1 (213) 740-2395
3620 S. Vermont Avenue, KAP 108 e-Mail:
Los Angeles, CA 90089-2532 On the Web: http://scf.usc.edu/~polunche
United States of America  

Education

8/2004 — 6/2009 Doctor of Philosophy Applied Mathematics
  Institution: University of Southern California,
Department of Mathematics,
Los Angeles, CA
  Emphasis: Mathematical Statistics, Sequential Quickest Change-Point Detection
  Thesis: Topics in Sequential Quickest Change-Point Detection
  Advisers: Remigijus Mikulevicius and Alexander G. Tartakovsky
8/2008 — 5/2009 Master of Science Mathematical Finance
  Institution: University of Southern California,
Department of Mathematics,
Los Angeles, CA
  Emphasis: Stochastic Modeling
8/2002 — 6/2004 Master of Science Applied Mathematics & Physics
  Institution: Moscow Institute of Physics and Technology,
Department of Electrical Engineering & Cybernetics,
Moscow, Russia
  Emphasis: Time Series Analysis; Image Processing & Visual Tracking
  Adviser:
8/1998 — 6/2002 Bachelor of Science Applied Mathematics & Physics
  Institution: Moscow Institute of Physics and Technology,
Department of Electrical Engineering & Cybernetics,
Moscow, Russia

Research Synopsis

My research interests are the general area of mathematical statistics and specifically sequential hypothesis testing, with particular emphasis on the problem of (optimal) sequential quickest change-point detection. The latter is considered under different settings: composite hypotheses where the pre- and/or post-change parameters are unknown, distributed multi-sensor setup where the information available for decision making is distributed across a set of (geographically separated) sensors, and multi-channel detection-identification where it is also desired to identify the channel(s) that ran out of control. This problem has an enormous spectrum of applications: quality control, financial markets, target tracking and detection — to name a few. Currently of interest is rapid intrusion detection in ultra high speed large-scale computer networks.

Areas of Professional Interest

Research Experience

9/2008 — Present Research Assistant University of Southern California,
Department of Mathematics,
Center for Applied Mathematical Sciences
  Grant Title: Optimal Changepoint Detection and Identification Algorithms with Applications
  Sponsoring Agency: National Science Foundation
  Principal Investigator (at USC): Prof. Alexander G. Tartakovsky
  Duties: Analysis and development of numerical methods for performance evaluation of various quickest change-point detection procedures such as Page's CUSUM, Shiryeav-Roberts and modifications thereof.
  Tools & Technologies used: MATLAB
10/2005 — 9/2008 Research Assistant University of Southern California,
Department of Mathematics,
Center for Applied Mathematical Sciences
  Grant Title: Bayesian Sensor Networks and Optimal Fusion Techniques in Multi-Sensor Distributed Systems with Applications to Counter Terrorism and Information Assurance
  Sponsoring Agency: Office of Naval Research
  Principal Investigator (at USC): Prof. Alexander G. Tartakovsky
  Duties: Development and performance evaluation of optimal and quasi-optimal change-point detection procedures in distributed sensor networks and applications to rapid intrusion detection in computer networks.
  Tools & Technologies used: MATLAB
6/2005 — Present Research Assistant University of Southern California,
Department of Mathematics,
Center for Applied Mathematical Sciences
  Grant Title: Spatio-Temporal Nonlinear Filtering with Applications to Information Assurance and Counter Terrorism
  Sponsoring Agency: Army Research Office
  Principal Investigator: Prof. Boris L. Rozovsky
  Co-Principal Investigator (at USC): Prof. Alexander G. Tartakovsky
  Duties: Development of efficient change-point detection procedures for composite hypotheses with applications to information assurance
  Tools & Technologies used: MATLAB (primarily Statistical Toolbox), ScaLAPACK, Goto BLAS, Message Passing Interface (MPI) 1 & 2, Intel Compiler for UNIX, GNU Scientific Library, USC High Performance Computing Parallel Cluster (Linux)
  URL: For more info click here
8/2004 — 12/2004 Research Assistant University of Southern California,
Department of Mathematics,
Center for Applied Mathematical Sciences
  Grant Title: Spatio-Temporal Image Restoration and Scene Stabilization Algorithm for High-Resolution Target Recognition
  Sponsoring Agency: Office of Naval Research
  Principal Investigator: Prof. Boris L. Rozovsky (at that time was affiliated with USC)
  Co-Principal Investigator: Prof. Alexander G. Tartakovsky
  Duties: Development of an Image Stabilization & Enhancement System incorporating adaptive spatial-temporal processing and an advanced jitter estimation/compensation/stabilization algorithm
  Tools & Technologies used: Microsoft Visual Studio, Intel Integrated Performance Primitives, Intel Open Computer Vision Library, Microsoft DirecX (DirectShow), Codejock XTreme ToolKit Pro, Assembler (x86)
  URL: For more info click here

Teaching Experience

1/2006 — 5/2006 Teaching Assistant University of Southern California,
Department of Mathematics
  Course Title: MATH 117 — Introduction to Mathematics for Business and Economics
  Course Outline: functions; graphs; polynomial and rational functions; exponential and logarithmic functions; matrices; systems of linear equations;
  Duties: leading discussion sessions (6 hours a week); holding office hours (3 hours a week); grading homework assignments; preparing and grading quizzes; grading exams; review sessions;
8/2005 — 12/2005 Teaching Assistant University of Southern California,
Department of Mathematics
  Course Title: MATH 126 — Calculus II
  Course Outline: trigonometric functions; applications of integration; techniques of integration; indeterminate forms; infinite series; Taylor series; polar coordinates;
  Duties: leading discussion sessions (6 hours a week); holding office hours (3 hours a week); grading homework assignments; preparing and grading quizzes; grading exams; review sessions;
1/2005 — 5/2005 Teaching Assistant University of Southern California,
Department of Mathematics
  Course Title: MATH 125 — Calculus I
  Course Outline: limits; continuity; derivatives and applications; antiderivatives; the fundamental theorem of calculus; exponential and logarithmic functions;
  Duties: leading discussion sessions (6 hours a week); holding office hours (3 hours a week); grading homework assignments; preparing and grading quizzes; grading exams; review sessions;
8/2003 — 6/2004 Assistant Lecturer Moscow Institute of Physics and Technology,
Department of Electrical Engineering & Cybernetics
Moscow, Russia
  Course Title: Linear Circuit Design (Lab)
  Course Outline: lumped circuit elements; network equations; zero-input and zero-state responses; sinusoidal steady-state analysis; impedance; resonance; network functions; power concepts; transformers; Laplace transforms;
  Duties: leading discussion sessions (10 hours a week); preparing and grading lab assignments; review sessions; lab equipment maintenance;

Awards and Honors

Publications

1. A.S. Polunchenko and A.G. Tartakovsky, On Optimality of the Shiryaev-Roberts Procedure for Detecting Changes in Distributions, (submitted to the Annals of Statistics in April), 2009. [arXiv]
2. G.V. Moustakides, A.S. Polunchenko and A.G. Tartakovsky, Design and Comparison of Shiryaev-Roberts- and CUSUM-Type Change-Point Detection Procedures, Proceedings of the 2nd International Workshop in Sequential Methodologies, University of Technology of Troyes, Troyes, France, 15 — 17 June 2009 (accepted, in press). [pdf]
3. C. Papadopoulos, A.S. Polunchenko and A.G. Tartakovsky, A Hybrid Approach to Efficient Detection of Attacks in Computer Networks, (in preparation, to be submitted to Computer Networks).
4. G.V. Moustakides, A.S. Polunchenko and A.G. Tartakovsky, Numerical Comparison of CUSUM and Shiryaev-Roberts Procedures for Detecting Changes in Distributions, Communications in Statistics: Theory and Methods, 2009 (accepted, in press). [pdf]
5. G.V. Moustakides, A.S. Polunchenko and A.G. Tartakovsky, A Numerical Approach to Performance Analysis of Quickest Change-Point Detection Procedures, Statistica Sinica, 2009 (accepted, in press). [pdf]
6. A.G. Tartakovsky and A.S. Polunchenko, Quickest Changepoint Detection in Distributed Multisensor Systems under Unknown Parameters, Proceedings of the 11th International Conference on Information Fusion, Hyatt Regency Hotel, Cologne, Germany, 30 June — 3 July 2008. [pdf]
7. A.G. Tartakovsky, M. Pollak and A.S. Polunchenko, Asymptotic Exponentiality of First Exit Times for Recurrent Markov Processes and Applications to Changepoint Detection, Proceedings of the 2008 International Workshop on Applied Probability, Compiégne, France, 7 — 10 July 2008. [pdf]
8. A.G. Tartakovsky and A.S. Polunchenko, Decentralized Quickest Change Detection in Distributed Sensor Systems with Applications to Information Assurance and Counter Terrorism, Proceedings of the 13th Annual Army Conference on Applied Statistics, Rice University, Houston, TX, 17 — 19 October 2007. [pdf]

Presentations & Seminar Talks

11/8/2007 Event: First Anual MURI Review Meeting
  Location: Emerald Room, Radisson Hotel, 3540 S. Figueroa St., Los Angeles, CA 90007
  Title: Hybrid Anomaly-Based Intrusion Detection System (IDS)
  Remarks: joint work with John Heidemann, Alexander G. Tartakovsky and Xue Cai
5/17/2007 Event: Special Interdisciplinary MURI Seminar
  Location: University of California at Los Angeles, Department of Mathematics
  Title: Efficiency of Binary Multi-Chart Change-Point Detection Procedures for Information Fusion in Distributed Multi-Sensor Networks: The Case of Composite Hypothesis
  Remarks: joint work with Alexander G. Tartakovsky

Professional Activities

References

Certifications

Computer Skills

Professional Societies

Languages


The University of Southern California does not screen or control the content on this website and thus does not guarantee the accuracy, integrity, or quality of such content. All content on this website is provided by and is the sole responsibility of the person from which such content originated, and such content does not necessarily reflect the opinions of the University administration or the Board of Trustees