Finite sample properties of the maximum likelihood estimator and of likelihood ratio tests in hidden Markov models. (English) Zbl 0986.62066
Summary: Hidden Markov models were successfully applied in various fields of time series analysis, especially for analyzing ion channel recordings. The maximum likelihood estimator (MLE) has recently been proven to be asymptotically normally distributed. Here, we investigate finite sample properties of the MLE and of different types of likelihood ratio tests (LRTs) by means of simulation studies. The MLE is shown to reach the asymptotic behavior within sample sizes that are common for various applications. Thus, reliable estimates and confidence intervals can be obtained. We give an approximative scaling function for the estimation error for finite samples, and investigate the power of different LRTs suitable for applications to ion channels, including tests for superimposed hidden Markov processes. Our results are applied to physiological sodium channel data.
MSC:
| 62M05 | Markov processes: estimation; hidden Markov models |
| 62P10 | Applications of statistics to biology and medical sciences; meta analysis |
Keywords:
ion channel analysis; asymptotic normality; superposition of Markov processes; power of test; sodium channel; time series; maximum likelihood estimator; finite sample properties; likelihood ratio tests; hidden Markov processesReferences:
| [1] | Albertsen, Biophys. J. 67 pp 1393– (1994) |
| [2] | Baum, Ann. Math. Stat. 41 pp 164– (1970) |
| [3] | Becker, Pfl. Arch. 426 pp 328– (1994) |
| [4] | Bickel, Ann. Stat. 26 pp 1614– (1998) |
| [5] | Chahine, Neuron 12 pp 281– (1994) |
| [6] | Chung, Math. Biosc. 133 pp 111– (1996) |
| [7] | Chung, Phil. Trans. R. Soc. Lond. B 329 pp 265– (1990) |
| [8] | and , 1974: Theoretical Statistics. Chapman & Hall. |
| [9] | , , and, 1998: Biological sequence analysis: Probabilistic models of proteins and nucleic acids. Cambridge University Press. · Zbl 0929.92010 |
| [10] | , and, 1995: Hidden Markov models: Estimation and control. Springer. · Zbl 0819.60045 |
| [11] | Fleischhauer, Pfl. Arch. 436 pp 757– (1998) |
| [12] | Fredkin, Proc. R. Soc. Lond. B 249 pp 125– (1992) |
| [13] | Giudici, Biometrics 56 pp 742– (2000) |
| [14] | Hamilton, J. Econometrics 45 pp 39– (1990) |
| [15] | 1994: Time Series Analysis. Princeton University Press. |
| [16] | Hoffmann, Cell 80 pp 681– (1995) |
| [17] | Horn, Biophys. J. 43 pp 207– (1983) |
| [18] | Horn, J. Gen. Physiol. 84 pp 505– (1984) |
| [19] | Kienker, Proc. R. Soc. Lond. B 236 pp 269– (1989) |
| [20] | Kim, J. Econometrics 60 pp 1– (1994) |
| [21] | Klein, Biometrics 53 pp 870– (1997) |
| [22] | Krogh, J. Mol. Biol. 235 pp 1501– (1994) |
| [23] | Leroux, Stoch. Proc. Appl. 40 pp 127– (1992) |
| [24] | Michalek, IEEE Trans. Sig. Proc. 47 pp 226– (1999) |
| [25] | Michalek, Eur. Biophys. J 28 pp 605– (1999) |
| [26] | Michalek, IEEE Trans. Sig. Proc. 48 pp 1537– (2000) |
| [27] | Milburn, Receptors Channels 3 pp 201– (1995) |
| [28] | Patlak, Physiol. Rev. 71 pp 1047– (1991) |
| [29] | , , and 1992: Numerical Recipes. Cambridge University Press. |
| [30] | Rabiner, Proc. IEEE 77 pp 257– (1989) |
| [31] | 1999: Bayesian Analysis of Hidden Markov Models for Single Ion Channel Records. Dissertation, University of Cambridge. |
| [32] | Timmer, Phys. Rev. E 55 pp 3306– (1997) |
| [33] | Venkataramanan, IEEE Trans. Sig. Proc. 46 pp 1901– (1998) |
| [34] | Wagner, Proc. R. Soc. Lond. B 266 pp 1919– (1999) |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
