This paper evaluates the application of the Minimum Classification Error (MCE) training to online-handwritten text recognition based on Hidden Markov Models. We describe an allograph-based, character level MCE training aimed at minimizing the character error rate while enabling flexibility in writing style. Experiments on a writer-independent discrete character recognition task covering all alp...

Discriminative training for hidden Markov models (HMMs) has been a central theme in speech recognition research for many years. One most popular technique is minimum classification error (MCE) training, with the objective function closely related to the empirical error rate and with the optimization method based traditionally on gradient descent. In this paper, we provide a new look at the MCE ...

We propose a deterministic annealing (DA) algorithm to design classifiers based on continuous observation hidden Markov models. The algorithm belongs to the class of minimum classification error (MCE) techniques that are known to outperform maximum likelihood (ML) design. Most MCE methods smooth the piecewise constant classification error cost to facilitate the use of local descent optimization...

INTRODUCTION AND OBJECTIVES It has been suggested that abnormal perfusion as derived from cardiovascular magnetic resonance imaging (CMR) is a transient dysfunction of microcirculation after myocardial infarction (MI) with TIMI 3 flow. We hypothesized that defects of myocardial perfusion may persist during the following months. METHODS Forty-seven patients with MI and sustained TIMI 3 flow un...

In this study, a new hidden Markov model that integrates generalized dynamic feature parameters into the model structure is developed and evaluated using maximum-likelihood (ML) and minimum-classification-error (MCE) pattern recognition approaches. In addition to the motivation of direct minimization of error rate, the MCE approach automatically eliminates the necessity of artificial constraint...

Minimum Classification Error (MCE) training is difficult to apply to language modeling due to inherent scarcity of training data (N-best lists). However, a whole-sentence exponential language model is particularly suitable for MCE training, because it can use a relatively small number of powerful features to capture global sentential phenomena. We review the model, discuss feature induction, fi...

This paper presents a prototype of a simulation model based on cellular automata (CA), and multicriteria evaluation (MCE) and integrated with GIS. Speci® cally, a method, analytical hierarchy process (AHP), of MCE is used here to derive behaviour-oriented rules of transition in CA. A t̀ight’ integration strategy is adopted, which means that the modules of MCE and CA are written in the C programm...

BACKGROUND The effects of high intensity interval training (HIIT) on inflammatory markers and endothelial function have been extensively shown. However, the acute effect of HIIT on platelet activation and function in patients with recent revascularization is unclear. OBJECTIVE The purpose of present study was to compare the responses of platelet activation (CD62P) and function (platelet aggre...

OBJECTIVES We aimed to ascertain whether triggered intravenous myocardial contrast echocardiography (MCE) can predict functional recovery in patients with acute myocardial infarction (AMI) and to determine the optimal triggering interval in this setting. BACKGROUND Detection of myocardial viability early after AMI has both therapeutic and prognostic implications. Myocardial contrast echocardi...

Compared with traditional gas-compression/expansion refrigeration, magnetic refrigeration based on magnetocaloric effect (MCE) exhibits the advantages of high energy efficiency and environment friendliness. Here, we created large MCE in RFeO3 (R = Tb or Tm) single crystals by the magnetization vector rotation of single crystal with strong magnetocrystalline anisotropy (MCA), rather than merely ...