A New Constant Memory Recursion for Hidden Markov Models

We develop the recursion for hidden Markov (HM) models proposed by Bartolucci and Besag (2002), and we show how it may be used to implement an estimation algorithm for these models that requires an amount of memory not depending on the length of the observed series of data. This recursion allows us to obtain the conditional distribution of the latent state at every occasion, given the previous state and the observed data. With respect to the estimation algorithm based on the well-known Baum-Welch recursions, which requires an amount of memory that increases with the sample size, the proposed algorithm also has the advantage of not requiring dummy renormalizations to avoid numerical problems. Moreover, it directly allows us to perform global decoding of the latent sequence of states, without the need of a Viterbi method and with a consistent reduction of the memory requirement with respect to the latter. The proposed approach is compared, in terms of computing time and memory requirement, with the algorithm based on the Baum-Welch recursions and with the so-called linear memory algorithm of Churbanov and Winters-Hilt. The comparison is also based on a series of simulations involving an HM model for continuous time-series data.

[1]  Francesco Bartolucci,et al.  Maximum likelihood estimation of a latent variable time‐series model , 2001 .

[2]  Josef Raviv,et al.  Decision making in Markov chains applied to the problem of pattern recognition , 1967, IEEE Trans. Inf. Theory.

[3]  H. Derin,et al.  A recursive algorithm for the Bayes solution of the smoothing problem , 1981 .

[4]  G. Lindgren.  Markov regime models for mixed distributions and switching regressions , 1978 .

[5]  Francesco Bartolucci,et al.  A recursive algorithm for Markov random fields , 2002 .

[6]  Richard Hughey,et al.  Reduced space sequence alignment , 1997, Comput. Appl. Biosci..

[7]  M. Bishop,et al.  Maximum likelihood alignment of DNA sequences. , 1986, Journal of molecular biology.

[8]  D. Haussler,et al.  Hidden Markov models in computational biology. Applications to protein modeling. , 1993, Journal of molecular biology.

[9]  István Miklós,et al.  A linear memory algorithm for Baum-Welch training , 2005, BMC Bioinformatics.

[10]  Gene Ott,et al.  Compact encoding of stationary Markov sources , 1967, IEEE Trans. Inf. Theory.

[11]  Sean R. Eddy,et al.  Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids , 1998 .

[12]  Francesco Bartolucci,et al.  Latent Markov Models for Longitudinal Data , 2012 .

[13]  Biing-Hwang Juang,et al.  Hidden Markov Models for Speech Recognition , 1991 .

[14]  Valeria De Fonzo,et al.  Hidden Markov Models in Bioinformatics , 2007 .

[15]  Francesco Bartolucci,et al.  Multidimensional Latent Markov Models in a Developmental Study of Inhibitory Control and Attentional Flexibility in Early Childhood , 2008, 0901.0024.

[16]  Francesco Bartolucci,et al.  Comment on the paper "On the memory complexity of the forward-backward algorithm, " by Khreich W., Granger E., Miri A., Sabourin, R , 2014, Pattern Recognit. Lett..

[17]  Hans Spada,et al.  Learning in Humans and Machines: Towards an Interdisciplinary Learning Science , 1995 .

[18]  Gary A. Churchill,et al.  Hidden Markov Chains and the Analysis of Genome Structure , 1992, Comput. Chem..

[19]  Richard Hughey,et al.  Reduced space hidden Markov model training , 1998, Bioinform..

[20]  G. Churchill.  Stochastic models for heterogeneous DNA sequences. , 1989, Bulletin of mathematical biology.

[21]  Stephen Winters-Hilt,et al.  Implementing EM and Viterbi algorithms for Hidden Markov Model in linear memory , 2007, BMC Bioinformatics.

[22]  Robert Sabourin,et al.  A survey of techniques for incremental learning of HMM parameters , 2012, Inf. Sci..

[23]  Robert Sabourin,et al.  On the memory complexity of the forward-backward algorithm , 2010, Pattern Recognit. Lett..

[24]  S. Taylor.  Asset Price Dynamics, Volatility, and Prediction , 2005 .

[25]  Geoffrey J. McLachlan,et al.  Finite Mixture Models , 2019, Annual Review of Statistics and Its Application.

[26]  S. L. Scott.  Bayesian Methods for Hidden Markov Models , 2002 .

[27]  L. Baum,et al.  A Maximization Technique Occurring in the Statistical Analysis of Probabilistic Functions of Markov Chains , 1970 .

[28]  Iain L. MacDonald,et al.  Some nonstandard stochastic volatility models and their estimation using structured hidden Markov models , 2012 .

[29]  James P. Hughes,et al.  Exact Computation of the Observed Information Matrix for Hidden Markov Models , 2002 .

[30]  Vasant Honavar,et al.  Learn++: an incremental learning algorithm for supervised neural networks , 2001, IEEE Trans. Syst. Man Cybern. Part C.

[31]  Andrew J. Viterbi,et al.  Error bounds for convolutional codes and an asymptotically optimum decoding algorithm , 1967, IEEE Trans. Inf. Theory.

[32]  Richard Hughey,et al.  Optimizing reduced-space sequence analysis , 2000, Bioinform..

[33]  Acidsby,et al.  Biological Sequence Analysis : Probabilistic Models of Proteins andNucleic , 2009 .

[34]  Stefano Ferilli,et al.  Avoiding Order Effects in Incremental Learning , 2005, AI*IA.