In this paper, we propose a new feature extraction method that combines both HMT-based denoising and weighted filter bank analysis for robust speech recognition. The proposed method is made up of two stages in cascade. The first stage is denoising process based on the wavelet domain Hidden Markov Tree model, and the second one is the filter bank analysis with weighting coefficients obtained fro...

This paper deals with Markov Decision Processes (MDPs) on Borel spaces with an infinite horizon and a discounted total cost. It will be considered a stochastic optimal control problem which arises by perturbing the transition law of a deterministic control problem, through an additive random noise term with coefficient epsilon. In the paper, we will analyze the behavior of the optimal solution ...

The Kalman filter is a recursive Best Linear Unbiased Estimator (BLUE) for a linear dynamic system with uncorrelated white process and measurement noises. It has been extended to the case where the noises are Markov and/or crosscorrelated for the same time instant. This paper presents optimal batch and semi-recursive filters and a suboptimal recursive filter for a linear discrete-time system wi...

This paper addresses the problem of robust speech recognition in noisy conditions in the framework of hidden Markov models (HMMs) and missing feature techniques. It presents a new statistical approach to detection and estimation of unreliable features based on a probabilistic measure and Gaussian mixture model (GMM). In the estimation process, the GMM is compensated using parameters of the stat...

Abstract. In this paper we consider an additive functional of an observable V (x) of a Markov jump process. We assume that the law of the expected jump time t(x) under the invariant probability measure π of the skeleton chain belongs to the domain of attraction of a subordinator. Then, the scaled limit of the functional is a Mittag-Leffler proces, provided that Ψ(x) := V (x)t(x) is square integ...

Traditional constraint-based and score-based methods for learning directed graphical models from continuous data have two significant limitations: (i) they require (in practice) assuming dependencies are linear with Gaussian noise; (ii) they cannot distinguish between Markov equivalent structures. More recent structure learning methods avoid both limitations by directly exploiting characteristi...

The distribution theory for reward functions on semi-Markov processes has been of interest since the early 1960s. The relevant asymptotic distribution theory has been satisfactorily developed. On the other hand, it has been noticed that it is difficult to find exact distribution results which lead to the effective computation of such distributions. Note that there is no satisfactory exact distr...

Central limit theorems and invariance principles are obtained for additive functionals of a stationary ergodic Markov chain, say Sn = g X1 + · · · + g Xn , where E g X1 =0 and E g X1 2 <∞. The conditions imposed restrict the moments of g and the growth of the conditional means E Sn X1 . No other restrictions on the dependence structure of the chain are required. When specialized to shift proces...