Training neural network acoustic models with sequencediscriminative criteria, such as state-level minimum Bayes risk (sMBR), been shown to produce large improvements in performance over cross-entropy. However, because they entail the processing of lattices, sequence criteria are much more computationally intensive than cross-entropy. We describe a distributed neural network training algorithm, ...

We describe the voice quality classification system that uses entropy and dynamism criteria as discrimination features. The main idea of this approach is that the input neural net is considered as an informational channel. Channel tuned to the certain type of information transmits it best of all according to the informational criterion. In our case a multilayer perceptron (MLP) emitted posterio...

Renyi’s entropy-based criterion has been proposed as an objective function for independent component analysis because of its relationship with Shannon’s entropy and its computational advantages in specific cases. These criteria were suggested based on “convincing” experiments. However, there is no theoretical proof that globally maximizing those functions would lead to separate the sources; act...

This paper provides a further look at uncertainty or information criteria used in the context of decision tree induction, and more generally of learning conditional class probability models. We show the high degree of similarity among two main families of criteria based respectively on the logarithmic SHANNON entropy function and the quadratic GINI index. We start by introducing a general famil...

We confront two integrability criteria for rational mappings. The first is the singularity confinement based on the requirement that every singu-larity, spontaneously appearing during the iteration of a mapping, disappear after some steps. The second recently proposed is the algebraic entropy criterion associated to the growth of the degree of the iterates. The algebraic entropy results confirm...

This paper is an extension of previous work which characterises soil behaviours using the grading entropy diagram. The present work looks at the piping process in granular soils, by considering some new data from flood-protection dikes. The piping process is divided into three parts here: particle movement at the micro scale to segregate free water; sand boil development (which is the initiatio...

Cross-entropy (CE)-based stopping criteria for turbo iterative decoding are known to outperform fixed-iteration stopping criteria at high signal-to-noise ratios (SNRs). While CE-based stopping criteria have a range of thresholds, a highvalue threshold for small frame sizes, and vice versa, should be used. It is difficult to advocate the value that can be categorized as either a small or large f...

This work relates to the solution of real multimedia documents indexing tasks. New approach to high-performance speech/music segmentation system creation is considered. This approach uses wavelet analysis and it is based not only on new distinctive features (entropy and dynamism) investigation, but it also takes into account information concerning language. Different experiments have shown the ...