On approximation of smoothing probabilities for hidden Markov models

We consider the smoothing probabilities of hidden Markov model (HMM). We show that under fairly general conditions for HMM, the exponential forgetting still holds, and the smoothing probabilities can be well approximated with the ones of double-sided HMM. This makes it possible to use ergodic theorems. As an application we consider the pointwise maximum a posteriori segmentation, and show that the corresponding risks converge.     

基于动态规划的高阶隐马氏模型推广的Viterbi算法

首先通过Hadar等价变换方法将高阶隐马氏模型转换为与之等价的一阶向量值隐马氏模型,然后利用动态规划原理建立了一阶向量值隐马氏模型的Viterbi算法,最后通过高阶隐马氏模型和一阶向量值隐马氏模型之间的等价关系建立了高阶隐马氏模型基于动态规划推广的Viterbi算法．研究结果在一定程度上推广了几乎所有隐马氏模型文献中所涉及到的解码问题的Viterbi算法,从而进一步丰富和发展了高阶隐马氏模型的算法理论．     

Generalisation of the Viterbi algorithm

The Viterbi algorithm, derived using dynamic programming techniques,is a maximum a posteriori (MAP) decoding method which was developedin the electrical engineering literature to be used in the analysisof hidden Markov models (HMMs). Given a particular HMM, theoriginal algorithm recovers the MAP state sequence underlyingany observation sequence generated from that model. This paperintroduces a generalization of the algorithm to recover, forarbitrary L, the top L most probable state sequences, with specialreference to its use in the area of automatic speech recognition.     

On the Accuracy of the MAP Inference in HMMs

In a hidden Markov model, the underlying Markov chain is usually unobserved. Often, the state path with maximum posterior probability (Viterbi path) is used as its estimate. Although having the biggest posterior probability, the Viterbi path can behave very atypically by passing states of low marginal posterior probability. To avoid such situations, the Viterbi path can be modified to bypass such states. In this article, an iterative procedure for improving the Viterbi path in such a way is proposed and studied. The iterative approach is compared with a simple batch approach where a number of states with low probability are all replaced at the same time. It can be seen that the iterative way of adjusting the Viterbi state path is more efficient and it has several advantages over the batch approach. The same iterative algorithm for improving the Viterbi path can be used when it is possible to reveal some hidden states and estimating the unobserved state sequence can be considered as an active learning task. The batch approach as well as the iterative approach are based on classification probabilities of the Viterbi path. Classification probabilities play an important role in determining a suitable value for the threshold parameter used in both algorithms. Therefore, properties of classification probabilities under different conditions on the model parameters are studied.     

Existence of infinite Viterbi path for pairwise Markov models

For hidden Markov models one of the most popular estimates of the hidden chain is the Viterbi path — the path maximizing the posterior probability. We consider a more general setting, called the pairwise Markov model, where the joint process consisting of finite-state hidden regime and observation process is assumed to be a Markov chain. We prove that under some conditions it is possible to extend the Viterbi path to infinity for almost every observation sequence which in turn enables to define an infinite Viterbi decoding of the observation process, called the Viterbi process. This is done by constructing a block of observations, called a barrier, which ensures that the Viterbi path goes through a given state whenever this block occurs in the observation sequence.     

基于二维真实隐马氏模型的手写体文字识别算法

近几年来，人们采用各种方法试图将1D隐马氏模型(HMM)^[2]推广到2D隐马氏模型。令人失望的是由于在建立合适的2D模型及其计算上的复杂度问题上存在困难，前面的尝试都没有得到一个真实的2DHMM．本文对于应用真实2D隐马氏模型(隐马氏网格随机场HMMRF)^[1,4]进行手写字符识别问题提出新的框架，针对文献[1]中的单点最优算法给出局部最优的译码算法。HMMRF模型是1D隐马氏模型到2D的扩展，能更好的描述字符的2D特性。HMMRF在字符识别中的应用具有两个相——学习相和译码相。在学习相和译码相中我们的最优标准是基于极大边缘后验概率的。不过，在涉及到2D模型中的计算问题时，对模型做出某些简单化的假设是必要的。本文用到的方法对于在合理的模型假设下解决手写字符识别问题呈现了很大的潜力。     

An S-Plus Implementation of Hidden Markov Models in Continuous Time

Abstract

Hidden Markov models (HMM) can be applied to the study of time varying unobserved categorical variables for which only indirect measurements are available. An S-Plus module to fit HMMs in continuous time to this type of longitudinal data is presented. Covariates affecting the transition intensities of the hidden Markov process or the conditional distribution of the measured response (given the hidden states of the process) are handled under a generalized regression framework. Users can provide C subroutines specifying the parameterization of the model to adapt the software to a wide variety of data types. HMM analysis using the S-Plus module is illustrated on a dataset from a prospective study of human papillomavirus infection in young women and on simulated data.     

A model for real-time failure prognosis based on hidden Markov model and belief rule base

As one of most important aspects of condition-based maintenance (CBM), failure prognosis has attracted an increasing attention with the growing demand for higher operational efficiency and safety in industrial systems. Currently there are no effective methods which can predict a hidden failure of a system real-time when there exist influences from the changes of environmental factors and there is no such an accurate mathematical model for the system prognosis due to its intrinsic complexity and operating in potentially uncertain environment. Therefore, this paper focuses on developing a new hidden Markov model (HMM) based method which can deal with the problem. Although an accurate model between environmental factors and a failure process is difficult to obtain, some expert knowledge can be collected and represented by a belief rule base (BRB) which is an expert system in fact. As such, combining the HMM with the BRB, a new prognosis model is proposed to predict the hidden failure real-time even when there are influences from the changes of environmental factors. In the proposed model, the HMM is used to capture the relationships between the hidden failure and monitored observations of a system. The BRB is used to model the relationships between the environmental factors and the transition probabilities among the hidden states of the system including the hidden failure, which is the main contribution of this paper. Moreover, a recursive algorithm for online updating the prognosis model is developed. An experimental case study is examined to demonstrate the implementation and potential applications of the proposed real-time failure prognosis method.     

Statistical inference for the nonparametric and semiparametric hidden Markov model via the composite likelihood approach

In this paper, we propose a new estimation method for a nonparametric hidden Markov model(HMM), in which both the emission model and the transition matrix are nonparametric, and a semiparametric HMM, in which the transition matrix is parametric while emission models are nonparametric. The estimation is based on a novel composite likelihood method, where the pairs of consecutive observations are treated as independent bivariate random variables. Therefore, the model is transformed into a mixture ...     

Filtering of a Multi-Dimension Stochastic Volatility Model

In this article, we study a stochastic volatility model for a class of risky assets. We assume that the volatilities of the assets are driven by a common state of economy, which is unobservable and represented by a hidden Markov chain. Under this hidden Markov model (HMM), we develop recursively computable filtering equations for certain functionals of the chain. Expectation maximization (EM) parameter estimation is then used. Applications to an optimal asset allocation problem with mean-variance utility are given.     

Computational issues in parameter estimation for stationary hidden Markov models

The parameters of a hidden Markov model (HMM) can be estimated by numerical maximization of the log-likelihood function or, more popularly, using the expectation–maximization (EM) algorithm. In its standard implementation the latter is unsuitable for fitting stationary hidden Markov models (HMMs). We show how it can be modified to achieve this. We propose a hybrid algorithm that is designed to combine the advantageous features of the two algorithms and compare the performance of the three algorithms using simulated data from a designed experiment, and a real data set. The properties investigated are speed of convergence, stability, dependence on initial values, different parameterizations. We also describe the results of an experiment to assess the true coverage probability of bootstrap-based confidence intervals for the parameters.     

Hidden semi-Markov model-based methodology for multi-sensor equipment health diagnosis and prognosis

This paper presents an integrated platform for multi-sensor equipment diagnosis and prognosis. This integrated framework is based on hidden semi-Markov model (HSMM). Unlike a state in a standard hidden Markov model (HMM), a state in an HSMM generates a segment of observations, as opposed to a single observation in the HMM. Therefore, HSMM structure has a temporal component compared to HMM. In this framework, states of HSMMs are used to represent the health status of a component. The duration of a health state is modeled by an explicit Gaussian probability function. The model parameters (i.e., initial state distribution, state transition probability matrix, observation probability matrix, and health-state duration probability distribution) are estimated through a modified forward–backward training algorithm. The re-estimation formulae for model parameters are derived. The trained HSMMs can be used to diagnose the health status of a component. Through parameter estimation of the health-state duration probability distribution and the proposed backward recursive equations, one can predict the useful remaining life of the component. To determine the “value” of each sensor information, discriminant function analysis is employed to adjust the weight or importance assigned to a sensor. Therefore, sensor fusion becomes possible in this HSMM based framework.     

隐马尔可夫模型在生物信息学中的应用

隐马尔可夫模型 ( HMM)是一个能够通过可观测的数据很好地捕捉真实空间统计性质的随机模型 ,该模型已成功地运用于语音识别 ,目前 HMM已开始应用于生物信息学 ( bioinformatics) ,已在生物序列分析中得到了广泛的应用 .本文首先介绍了 HMM的基本结构 ,然后着重讨论了 HMM在 DNA序列的多重比对 ,基因发现等生物序列分析中的应用     

Filter‐based portfolio strategies in an HMM setting with varying correlation parametrizations

We consider portfolio optimization in a regime‐switching market. The assets of the portfolio are modeled through a hidden Markov model (HMM) in discrete time, where drift and volatility of the single assets are allowed to switch between different states. We consider different parametrizations of the involved asset covariances: statewise uncorrelated assets (though linked through the common Markov chain), assets correlated in a state‐independent way, and assets where the correlation varies from state to state. As a benchmark, we also consider a model without regime switches. We utilize a filter‐based expectation‐maximization (EM) algorithm to obtain optimal parameter estimates within this multivariate HMM and present parameter estimators in all three HMM settings. We discuss the impact of these different models on the performance of several portfolio strategies. Our findings show that for simulated returns, our strategies in many settings outperform naïve investment strategies, like the equal weights strategy. Information criteria can be used to detect the best model for estimation as well as for portfolio optimization. A second study using real data confirms these findings.     

On Adjusted Viterbi Training

The EM algorithm is a principal tool for parameter estimation in the hidden Markov models, where its efficient implementation is known as the Baum–Welch algorithm. This paper is however motivated by applications where EM is replaced by Viterbi training, or extraction (VT), also known as the Baum–Viterbi algorithm. VT is computationally less intensive and more stable, and has more of an intuitive appeal. However, VT estimators are also biased and inconsistent. Recently, we have proposed elsewhere the adjusted Viterbi training (VA), a new method to alleviate the above imprecision of the VT estimators while preserving the computational advantages of the baseline VT algorithm. The key difference between VA and VT is that asymptotically, the true parameter values are a fixed point of VA (and EM), but not of VT. We have previously studied VA for a special case of Gaussian mixtures, including simulations to illustrate its improved performance. The present work proves the asymptotic fixed point property of VA for general hidden Markov models. J. Lember is supported by Estonian Science Foundation Grant 5694.     

Maximum likelihood estimation for stochastic volatility in mean models with heavy‐tailed distributions

In this article, we introduce a likelihood‐based estimation method for the stochastic volatility in mean (SVM) model with scale mixtures of normal (SMN) distributions. Our estimation method is based on the fact that the powerful hidden Markov model (HMM) machinery can be applied in order to evaluate an arbitrarily accurate approximation of the likelihood of an SVM model with SMN distributions. Likelihood‐based estimation of the parameters of stochastic volatility models, in general, and SVM models with SMN distributions, in particular, is usually regarded as challenging as the likelihood is a high‐dimensional multiple integral. However, the HMM approximation, which is very easy to implement, makes numerical maximum of the likelihood feasible and leads to simple formulae for forecast distributions, for computing appropriately defined residuals, and for decoding, that is, estimating the volatility of the process. Copyright © 2017 John Wiley & Sons, Ltd.     

Maximum likelihood estimation for general hidden semi-Markov processes with backward recurrence time dependence

This paper concerns the study of asymptotic properties of the maximum likelihood estimator (MLE) for the general hidden semi-Markov model (HSMM) with backward recurrence time dependence. By transforming the general HSMM into a general hidden Markov model, we prove that under some regularity conditions, the MLE is strongly consistent and asymptotically normal. We also provide useful expressions for asymptotic covariance matrices, involving the MLE of the conditional sojourn times and the embedded Markov chain of the hidden semi-Markov chain. Bibliography: 17 titles.     

Interactive hidden Markov models and their applications

^** Email: wching{at}hkusua.hku.hk In this paper, we propose an Interactive hidden Markov model(IHMM). In a traditional HMM, the observable states are affecteddirectly by the hidden states, but not vice versa. In the proposedIHMM, the transitions of hidden states depend on the observablestates. We also develop an efficient estimation method for themodel parameters. Numerical examples on the sales demand dataand economic data are given to demonstrate the applicabilityof the model.     

HMM based scenario generation for an investment optimisation problem

The Geometric Brownian motion (GBM) is a standard method for modelling financial time series. An important criticism of this method is that the parameters of the GBM are assumed to be constants; due to this fact, important features of the time series, like extreme behaviour or volatility clustering cannot be captured. We propose an approach by which the parameters of the GBM are able to switch between regimes, more precisely they are governed by a hidden Markov chain. Thus, we model the financial time series via a hidden Markov model (HMM) with a GBM in each state. Using this approach, we generate scenarios for a financial portfolio optimisation problem in which the portfolio CVaR is minimised. Numerical results are presented.     

Efficient sensitivity analysis in hidden markov models

Sensitivity analysis in hidden Markov models (HMMs) is usually performed by means of a perturbation analysis where a small change is applied to the model parameters, upon which the output of interest is re-computed. Recently it was shown that a simple mathematical function describes the relation between HMM parameters and an output probability of interest; this result was established by representing the HMM as a (dynamic) Bayesian network. To determine this sensitivity function, it was suggested to employ existing Bayesian network algorithms. Up till now, however, no special purpose algorithms for establishing sensitivity functions for HMMs existed. In this paper we discuss the drawbacks of computing HMM sensitivity functions, building only upon existing algorithms. We then present a new and efficient algorithm, which is specially tailored for determining sensitivity functions in HMMs.