Stilian A. StoevAssociate Professor,Department of Statistics University of Michigan, Ann Arbor |
We are what we repeatedly do. Excellence, therefore, is not an act,
but a habit. -- Aristotle
Welcome
You can find out more about my research, publications, teaching, and download some free Matlab, Octave, and R code by clicking on the above tabs.
Contact Information
Mailing Address:
439 W. Hall, 1085 S. University
Ann Arbor, MI 48109-1107, USA
Office: 445C West Hall
Phone: +1.734.763.9294
Email: sstoev.at.umich.edu
Research Interests
Time series, long-range dependence, heavy tails and extremes. Spatial and spatio-temporal models. Applications to Internet traffic modeling, computer science, and finance.
Education
(1999-2005) Ph.D. in Mathematics and Statistics from Boston University with Professor Murad S. Taqqu.
(1994-1998) Master's in Mathematics Sofia University with Professor Dimitar L. Vandev.
(1989-1994) National School for Natural and Mathematical Sciences, "Acad. L. Chakalov", Sofia, Bulgaria: NPMG.
Vita
CV in PDF.
My general research interests are in applied probability and statistics: time series analysis, long-range dependent and heavy-tailed models, self-similarity, stable processes, max-stable processes and their applications to Internet traffic modeling, computer science and finance.
Probability
My most recent work focuses on extreme value theory. I am interested in the represenations, ergodic properties, structure and classification of max-stable processes and random fields. Exciting applications of max-stable processes and random fields arise in many fields where extremes and maxima are of interest.
Statistics and Applications
Internet traffic modeling, streaming data, estimation of the tail exponent and the extremal index. Prediction in max-stable models. Spatial statisitcs and wavelets.
Computing
Efficient simulation of stochstic processes: FARIMA time series, fractional Brownian motion, multifractional Brownian motion, linear fractional stable motions, max-stable processes and random fields. Computational prediction in max-stable random fields.
For more details and free Matlab and R code, click on the Software tab above.
Undergraduate Courses
This course provides a first look at probability and statistics for undergraduate students in the engineering and computer science fields. It assumes moderate calculus background and covers equal amounts of probability and statisitcs.
Calculus-based introduction to basic probability for undergraduate students interested in statistics, mathematics, actuarial science, engineering, or other quantitative fields.
Graduate Courses
This course provides a rigorous introduction to Probability Theory. The required measure theoretic background is developed from scratch. Then the concepts of independence, integration, and expectation are introduced. Special attention is payed on various modes of convergence that are ubiquitous in Probability and Statistics. The laws of large numbers and the central limit theorem are proved. The course concludes with a careful treatment of conditional expectation via the Radon-Nikodym Theorem and basic results on martingales.
The course provides background on topics from advanced linear algebra, real analysis, and measure theory, indispensable for the graduate studies in statistics.
Special topics course for graduate students reviewing recent advances on theory and applications involving dependent data with emphasis on Internet traffic modeling.
Selected Papers
Recent Preprints
[1] S. Stoev, V. Pipiras & M.S. Taqqu (2002) ``Estimation of the self-similarity parameter in linear fractional stable motion''. Signal Processing, {82} (2002), pp 1873-1901.
[2] S. Stoev & M.S. Taqqu (2004) ``Simulation methods for linear fractional stable motion and FARIMA using the Fast Fourier Transform'', Fractals, 12(1) (2004), pp 95-121.
[3] S. Stoev & M.S. Taqqu (2004) ``Stochastic properties of the linear multifractional stable motion'', Advances in Applied Probability, 36 (2004), pp 1085-1115.
[4] S. Stoev & M.S. Taqqu (2005) ``Asymptotic self-similarity and wavelet estimation for long-range dependent FARIMA time series with stable innovations'', Journal of Time Series Analysis, 26(2) (2005), pp 211-249.
[5] S. Stoev & M.S. Taqqu (2005) ``Path properties of the linear multifractional stable motion'', Fractals, 13(2) (2005), pp 157-178.
[6] S. Stoev, M.S. Taqqu, C. Park & J.S. Marron (2005) ``On the wavelet spectrum diagnostic for Hurst parameter estimation in the analysis of Internet traffic'', Computer Networks, 48 (2005), pp 423-445.
[7] S. Stoev & M.S. Taqqu (2005) ``Weak convergence to the tangent process of the linear multifractional stable motion'', PLISKA - Studia Mathematica Bulgarica, 17 (2005), pp 271-294.
[8] S. Stoev & M.S. Taqqu (2006) ``How rich is the class of multifractional Brownian motions?'', Stochastic Processes and Their Applications, 116(2) (2006), pp 200-221.
[9] S. Stoev, M.S. Taqqu, C. Park, G. Michailidis & J.S. Marron (2006) ``LASS: a tool for the local analysis of self-similarity'', Computational Statistics and Data Analysis, 50 (2006), pp 2447-2471.
[10] S. Stoev & M.S. Taqqu (2005) ``Extremal stochastic integrals: a parallel between max-stable processes and &alpha -stable processes'', Extremes, 8 (2005), pp 237-266.
[11] C. Park, F. Godtliebsen, S. Stoev, M.S. Taqqu & J.S. Marron (2007) ``Visualization and inference based on wavelet coefficients SiZer and SiNos''. Journal of Computational and Graphical Statistics , 51 (2007), pp 5994-6012.
[12] S. Stoev & M.S. Taqqu (2007) ``Limit theorems for sums of heavy-tailed terms with random dependent weights''. Methodology and Computing in Applied Probability, 9 (2007), pp 55-87.
[13] S. Stoev & M.S. Taqqu (2007) ``Limit theorems for maxima of heavy-tailed terms with random dependent weights'', PLISKA - Studia Mathematica Bulgarica, 18 (2007), pp 361-378.
[14] P.-L. Conti, L. De Giovanni, S. Stoev & M.S. Taqqu (2008) ``Confidence intervals for the long memory parameter based on wavelets and resampling''. Statistica Sinica, 18(2), pp 559-579.
[15] S. Stoev (2008) ``On the ergodicity and mixing of max-stable processes''. Stochastic Processes and their Applications, 118(9), pp 1679-1705.
[16] M. Meerschaert & S. Stoev (2009) ``Extremal limit theorems for observations separated by random waiting times'', Journal of Statistical Planning and Inference, 139(7) (2009), pp 2175-2188.
[17] Stoev, S. and Michailidis, G. (2009) ``On the estimation of the heavy-tail exponent in time series using the max-spectrum'', Applied Stochastic Models in Business and Industry.
[18] Hamidieh, K, Stoev, S., & Michailidis, G. (2009) ``On the estimation of the extremal index based on scaling and resampling'', Journal of Computational and Graphical Statistics, 18(3) (2009) pp 731-755.
[19] Y. Wang & S. Stoev (2009) ``On the association of sum- and max-stable processes'', Statistics and Probability Letters, 80 (2010) pp 480-488.
[20] Y. Wang & S. Stoev (2010) ``On the structure and representations of max-stable processes'', Advances in Applied Probability, 42(3), pp 855-877.
[21] A. Ruzmaikin, J. Feynman, & S. Stoev (2011), ``Distribution and clustering of fast coronal mass ejections'', Journal of Geophysical Research, 116.
[22] S. Stoev, G. Michailidis & M.S. Taqqu (2011) ``Estimating heavy-tail exponents through max self-similarity'', IEEE Transactions on Information Theory, 57(3), pp 1615-1635.
Preprints and Technical Reports Top
[1] S. Stoev & G. Michailidis (2006) On the estimation of the heavy-tail exponent in time series using the max-spectrum. Department of Statistics, the University of Michigan, Technical Report 447: arXiv | PDF.
[2] S. Stoev & M.S. Taqqu (2006) Max-stable sketches: estimation of L_{&alpha}-norms, dominance norms and point queries for non-negative signals Department of Statistics, the University of Michigan, Technical Report 433: arXiv | PDF.
[3] S. Stoev, G. Michailidis & M.S. Taqqu (2007) Estimating heavy-tail exponents through max self-similarity. Department of Statistics, the University of Michigan, Technical Report 447: arXiv | PDF.
[4] K. Hamidieh, S. Stoev & G. Michailidis (2007) On the estimation of the extremal index based on scaling and resampling, Department of Statistics, the University of Michigan, Technical Report 462: arXiv | PDF.
[5] Y. Wang & S. Stoev (2009) On the Structure and Representations of Max-Stable Processes, Department of Statistics, the University of Michigan, Technical Report 487: arXiv| PDF.
[6] K. Hamidieh, S. Stoev & G. Michailidis (2009) Intensity Based Estimation of Value-at-Risk , submitted to CSDA.
[7] Y. Wang, P. Roy & S. Stoev (2009) Ergodic properties of sum- and max- stable stationary random fields via null and positive group actions, Department of Statistics, the University of Michigan, Technical Report 501: arXiv | PDF.
[8] S. Stoev, G. Michailidis & J. Vaughan (2009) Global Modeling and Prediction of Computer Network Traffic, Department of Statistics, the University of Michigan, Technical Report 490: arXiv | PDF.
[9] A. Ruzmaikin, J. Feynman, and S. Stoev (2010), Distribution and Clustering of Fast Coronal Mass Ejections, submitted to Geophysical Research Letters.
[10] J. Vaughan, S. Stoev, & G. Michailidis (2010), Network-wide Statistical Modeling and Prediction of Computer Traffic, Department of Statistics, the University of Michigan, Technical Report 501: arXiv | PDF.
[11] Y. Wang & S. Stoev (2010) Conditional Sampling for Max-Stable Random Fields, Department of Statistics, the University of Michigan, Technical Report 509: arXiv | PDF | Software.
[12] S. Stoev (2011) Functional Limit Theorems for Maxima in Hölder Spaces, Department of Statistics, the University of Michigan, Technical Report 518: PDF
[13] Y. Wang, S. Stoev & P. Roy (2011) Decomposability for Stable Processes, Department of Statistics, the University of Michigan, Technical Report 520: arXiv | PDF
Book Chapters and Conference Proceedings Top
[1] S. Stoev & M.S. Taqqu (2003) ``Wavelet estimation of the Hurst parameter in stable processes''In: Processes with Long Range Correlations: Theory and Applications, G. Rangarajan and M. Ding editors, Springer Verlag, Berlin, 2003, Lecture Notes in Physics No. 621, pp 61-87.
[2] J.-M. Bardet, G. Lang, G. Oppenheim, A. Philippe, M.S. Taqqu & S. Stoev (2003) ``Semi-parametric estimation of the long-range dependence parameter: A survey''. In: Theory and Applications of Long-range Dependence, P. Doukhan, G. Oppenheim, and M.S. Taqqu, editors, Birkhauser, Boston, 2003, pp 579-623.
[3] S. Stoev, M. Hadjieleftheiou, G. Kollios & M.S. Taqqu (2007) ``Norm, point, and distance estimation over multiple signals using max-stable distributions''. in Proceedings of the 23rd International Conference on Data Engineering ICDE, Istanbul, Turkey, April 2007. Acceptance rate: 18.5%.
[4] S. Stoev (2010) ``Max-Stable Processes: Representations, Ergodic Properties and Statistical Applications''. In: Dependence, with Applications in Statistics and Econometrics, P. Doukhan, G. Lang, D. Surgailis and G. Teyssiere, editors, Springer, New York, Lecture Notes in Statistics.
[5] S. Stoev, G. Michailidis, and J. Vaughan (2010) ``On Global Modeling of Network Traffic'', INFOCOM 2010, The 29th Conference on Computer Communications, San Diego, California, March 2010. Acceptance rate: 17%.
Recent Presentations
Current
Topic: Statistical modeling and inference of spatial extremes.
Topic: Shape restricted inference for dependent data.
Past
Thesis:
Inference for neuronal networks using temporal and graphical
models.
Currently: Senior Bio-statistician Novartis Oncology in
Hyderabad, India.
Thesis:
Problems in spatio-temporal modeling, kriging, and prediction of
computer network traffic.
Currently: Assistant Professor at Quinnipiac University in
Hamden, Connecticut.
Thesis: Topics on max-stable processes and the Central Limit
Theorem
Winner of the
ProQuest Distinguished Dissertation Award of the Rackham Graduate School.
Currently: Assistant Professor at University of Cincinnati, Ohio.
Thesis: Topics in Statistical Modeling and Estimation
of Extremes and Their Dependence.
Currently: Assistant Professor at California State University,
Fullerton.
Thesis: Modeling of Extremes with Applications to Insurance Claims Data.
Long Range Dependence
Extremes
Empirical findings
Thoughts of others
"I've seen the nations rise and fall
I've heard their stories, heard them all
But love's the only engine of survival..."
Useful Links
Wikipedia | MathSciNet | Jstor | .
Maps, Directions, & Campus Info
gs -dBATCH -dNOPAUSE -q -sDEVICE=pdfwrite -sOutputFile=finished.pdf file1.pdf file2.pdf file3.pdf
wrapped in a self-explanatory bash script pdfmerge. For more details, see the actual post.
Some Pictures Graybill VIII/EVA 2009 | More to come ...