A normalized statistical metric space for hidden Markov models

Chen Lu, Jason M. Schwier, Ryan M. Craven, Lu Yu, Richard R. Brooks, Christopher Griffin

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

In this paper, we present a normalized statistical metric space for hidden Markov models (HMMs). HMMs are widely used to model real-world systems. Like graph matching, some previous approaches compare HMMs by evaluating the correspondence, or goodness of match, between every pair of states, concentrating on the structure of the models instead of the statistics of the process being observed. To remedy this, we present a new metric space that compares the statistics of HMMs within a given level of statistical significance. Compared with the Kullback-Leibler divergence, which is another widely used approach for measuring model similarity, our approach is a true metric, can always return an appropriate distance value, and provides a confidence measure on the metric value. Experimental results are given for a sample application, which quantify the similarity of HMMs of network traffic in the Tor anonymization system. This application is interesting since it considers models extracted from a system that is intentionally trying to obfuscate its internal workings. In the conclusion, we discuss applications in less-challenging domains, such as data mining.

Original languageEnglish (US)
Pages (from-to)806-819
Number of pages14
JournalIEEE Transactions on Cybernetics
Volume43
Issue number3
DOIs
StatePublished - Jun 1 2013

Fingerprint

Hidden Markov models
Statistics
Data mining

All Science Journal Classification (ASJC) codes

  • Software
  • Control and Systems Engineering
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Lu, Chen ; Schwier, Jason M. ; Craven, Ryan M. ; Yu, Lu ; Brooks, Richard R. ; Griffin, Christopher. / A normalized statistical metric space for hidden Markov models. In: IEEE Transactions on Cybernetics. 2013 ; Vol. 43, No. 3. pp. 806-819.
@article{a3fc9006ebd7476698d33f84b3b1cc6f,
title = "A normalized statistical metric space for hidden Markov models",
abstract = "In this paper, we present a normalized statistical metric space for hidden Markov models (HMMs). HMMs are widely used to model real-world systems. Like graph matching, some previous approaches compare HMMs by evaluating the correspondence, or goodness of match, between every pair of states, concentrating on the structure of the models instead of the statistics of the process being observed. To remedy this, we present a new metric space that compares the statistics of HMMs within a given level of statistical significance. Compared with the Kullback-Leibler divergence, which is another widely used approach for measuring model similarity, our approach is a true metric, can always return an appropriate distance value, and provides a confidence measure on the metric value. Experimental results are given for a sample application, which quantify the similarity of HMMs of network traffic in the Tor anonymization system. This application is interesting since it considers models extracted from a system that is intentionally trying to obfuscate its internal workings. In the conclusion, we discuss applications in less-challenging domains, such as data mining.",
author = "Chen Lu and Schwier, {Jason M.} and Craven, {Ryan M.} and Lu Yu and Brooks, {Richard R.} and Christopher Griffin",
year = "2013",
month = "6",
day = "1",
doi = "10.1109/TSMCB.2012.2216872",
language = "English (US)",
volume = "43",
pages = "806--819",
journal = "IEEE Transactions on Cybernetics",
issn = "2168-2267",
publisher = "IEEE Advancing Technology for Humanity",
number = "3",

}

A normalized statistical metric space for hidden Markov models. / Lu, Chen; Schwier, Jason M.; Craven, Ryan M.; Yu, Lu; Brooks, Richard R.; Griffin, Christopher.

In: IEEE Transactions on Cybernetics, Vol. 43, No. 3, 01.06.2013, p. 806-819.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A normalized statistical metric space for hidden Markov models

AU - Lu, Chen

AU - Schwier, Jason M.

AU - Craven, Ryan M.

AU - Yu, Lu

AU - Brooks, Richard R.

AU - Griffin, Christopher

PY - 2013/6/1

Y1 - 2013/6/1

N2 - In this paper, we present a normalized statistical metric space for hidden Markov models (HMMs). HMMs are widely used to model real-world systems. Like graph matching, some previous approaches compare HMMs by evaluating the correspondence, or goodness of match, between every pair of states, concentrating on the structure of the models instead of the statistics of the process being observed. To remedy this, we present a new metric space that compares the statistics of HMMs within a given level of statistical significance. Compared with the Kullback-Leibler divergence, which is another widely used approach for measuring model similarity, our approach is a true metric, can always return an appropriate distance value, and provides a confidence measure on the metric value. Experimental results are given for a sample application, which quantify the similarity of HMMs of network traffic in the Tor anonymization system. This application is interesting since it considers models extracted from a system that is intentionally trying to obfuscate its internal workings. In the conclusion, we discuss applications in less-challenging domains, such as data mining.

AB - In this paper, we present a normalized statistical metric space for hidden Markov models (HMMs). HMMs are widely used to model real-world systems. Like graph matching, some previous approaches compare HMMs by evaluating the correspondence, or goodness of match, between every pair of states, concentrating on the structure of the models instead of the statistics of the process being observed. To remedy this, we present a new metric space that compares the statistics of HMMs within a given level of statistical significance. Compared with the Kullback-Leibler divergence, which is another widely used approach for measuring model similarity, our approach is a true metric, can always return an appropriate distance value, and provides a confidence measure on the metric value. Experimental results are given for a sample application, which quantify the similarity of HMMs of network traffic in the Tor anonymization system. This application is interesting since it considers models extracted from a system that is intentionally trying to obfuscate its internal workings. In the conclusion, we discuss applications in less-challenging domains, such as data mining.

UR - http://www.scopus.com/inward/record.url?scp=84890426084&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890426084&partnerID=8YFLogxK

U2 - 10.1109/TSMCB.2012.2216872

DO - 10.1109/TSMCB.2012.2216872

M3 - Article

VL - 43

SP - 806

EP - 819

JO - IEEE Transactions on Cybernetics

JF - IEEE Transactions on Cybernetics

SN - 2168-2267

IS - 3

ER -