Source Estimation in Time Series and the Surprising Resilience of HMMs

@article{Kozdoba2018SourceEI,
  title={Source Estimation in Time Series and the Surprising Resilience of HMMs},
  author={Mark Kozdoba and Shie Mannor},
  journal={IEEE Transactions on Information Theory},
  year={2018},
  volume={64},
  pages={5555-5569}
}
Suppose that we are given a time series where consecutive samples are believed to come from a probabilistic source, that the source changes from time to time and that the total number of sources is fixed. Our objective is to estimate the distributions of the sources. A standard approach to this problem is to model the data as a hidden Markov model (HMM). However, since the data often lacks the Markov or the stationarity properties of an HMM, one can ask whether this approach is still suitable… CONTINUE READING

Topics from this paper.

References

Publications referenced by this paper.
SHOWING 1-10 OF 35 REFERENCES

and S

J. P. Hughes, P. Guttorp
  • P. Charles, “A non-homogeneous hidden Markov model for precipitation occurrence,” J. Roy. Statist. Soc., C (Appl. Statist.), vol. 48, no. 1, pp. 15–30
  • 1999
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

A method of moments for mixture models and hidden Markov models

T. M. Cover Csiszár, B.-S. Choi
  • Proc . COLT
  • 2012

and D

M. Kritzman, S. Page
  • Turkington, “Regime shifts: Implications for dynamic strategies,” Financial Analysts J., vol. 68, no. 3, pp. 22–39
  • 2012
VIEW 2 EXCERPTS

and S

A. Anandkumar, D. Hsu
  • M. Kakade, “A method of moments for mixture models and hidden Markov models,” in Proc. COLT
  • 2012

Consistency of the maximum likelihood estimator for general hidden Markov models

E. Moulines Douc, J. Olsson, R. van Handel
  • Ann . Statist .
  • 2011

and J

L. Huang, A. D. Joseph, B. Nelson, B. I. Rubinstein
  • D. Tygar, “Adversarial machine learning,” in Proc. 4th ACM Workshop Secur. Artif. Intell.
  • 2011

Similar Papers

Loading similar papers…