It is well known that speaker identification yields very high performance in a neutral talking environment; on the other hand, the performance has been sharply declined in a shouted talking environment. This work aims at proposing, implementing, and evaluating novel Third-Order Circular Suprasegmental Hidden Markov Models (CSPHMM3s) to improve the low performance of text-independent speaker ide...

Today’s state-of-the-art speech recognition systems typically use continuous density hidden Markov models with mixture of Gaussian distributions. Such speech recognition systems have problems; they require too much memory to run, and are too slow for large vocabulary applications. Two approaches are proposed for the design of compact acoustic models, namely, subspace distribution clustering hid...

We prove that under a mild positivity assumption the entropy rate of a hidden Markov chain varies analytically as a function of the underlying Markov chain parameters. We give examples to show how this can fail in some cases. And we study two natural special classes of hidden Markov chains in more detail: binary hidden Markov chains with an unambiguous symbol and binary Markov chains corrupted ...

We prove that under a mild positivity assumption the entropy rate of a hidden Markov chain varies analytically as a function of the underlying Markov chain parameters. We give examples to show how this can fail in some cases. And we study two natural special classes of hidden Markov chains in more detail: binary hidden Markov chains with an unambiguous symbol and binary Markov chains corrupted ...

Multi-state Markov models have proven useful in many concrete situations and have had successful applications, notably in analysing disease history data. Continuous-time Markov models with three transient states are considered for the study of asthma control evolution in a cohort of 406 patients with persistent asthma. Firstly, the model is considered in the homogeneous case. Covariates are tak...

Inspired by the recent work of Daubechies and Lagarias on a set of matrices with convergent innnite products, we establish a new geometric approach to the classical problem of (weakly) ergodic non-homogeneous Markov chains. The existing key inequalities (related to the Hajnal inequality) in the literature are uniied in this geometric picture. A more general inequality is established. Important ...

Let us consider a homogeneous Markov chain with discrete time and with a finite set of states are nonrecurrent. The goal of this work is to study frequencies of trajectories in this chain, i.e., " words " composed of symbols E 1 ,. .. , E n ending with the " space " E 0. Let us order words according to their probabilities ; denote by p(t) the probability of the tth word in this list. In this pa...

Hidden Markov Model on First Order Differences for Automatic Detection of Nucleosome Positions” Pei Fen Kuan1, Dana Huebert2, Audrey Gasch3, Sündüz Keleş1,4∗ 1Department of Statistics, University of Wisconsin, Madison, WI 53706. 2Department of Cellular and Molecular Biology, University of Wisconsin, Madison, WI 53706. 3Department of Genetics, University of Wisconsin, Madison, WI 53706. 4Departm...

A number of software reliability models have been proposed for assessing the reliability of a software system. In this paper, we discuss the time-domain and data-domain approaches to software reliability model-ing, and classify the previously reported models into these two classes based on their underlying assumptions. The data-domain models are further classiied into fault-seeding and input do...