Fast Minimum Training Error Discretization
نویسندگان
چکیده
منابع مشابه
A Fast Discriminative Training Algorithm for Minimum Classification Error
In this paper a new algorithm is proposed for fast discriminative training of hidden Markov models (HMMs) based on minimum classification error (MCE). The algorithm is able to train acoustic models in a few iterations, thus overcoming the slow training speed typical of discriminative training methods based on gradient-descendent. The algorithm tries to cancel the gradient of the objective funct...
متن کاملA Receiver Structure for Frequency-Flat Time-Varying Rayleigh Channels and Performance Analysis
This paper proposes a wavelet based receiver structure for frequency-flat time-varying Rayleigh channels, consisting of a receiver front-end followed by a Maximum A-Posteriori (MAP) detector. Discretization of the received continuous time signal using filter banks is an essential stage in the front-end part, where the Fast Haar Transform (FHT) is used to reduce complexity. Analysis of our recei...
متن کاملError-Based and Entropy-Based Discretization of Continuous Features
We present a comparison of error-based and entropybased methods for discretization of continuous features. Our study includes both an extensive empirical comparison as well as an analysis of scenarios where error minimization may be an inappropriate discretization criterion. We present a discretization method based on the C4.5 decision tree algorithm and compare it to an existing entropy-based ...
متن کاملVine Parsing and Minimum Risk Reranking for Speed and Precision
We describe our entry in the CoNLL-X shared task. The system consists of three phases: a probabilistic vine parser (Eisner and N. Smith, 2005) that produces unlabeled dependency trees, a probabilistic relation-labeling model, and a discriminative minimum risk reranker (D. Smith and Eisner, 2006). The system is designed for fast training and decoding and for high precision. We describe sources o...
متن کاملA New Technique for the Calculation of Colliding Vortex Rings
The present study involves a novel computational technique, regarding simultaneous use of the pseudo particle method, Poisson integral method and a special-purpose computer originally designed for molecular dynamics simulations (MDGRAPE-3). In the present calculations, the dynamics of two colliding vortex rings have been studied using the vortex method. The present acceleration technique allows...
متن کامل