For extended periods of time, sequence generation models rely on beam search as the decoding algorithm. However, the performance of beam search degrades when the model is over-confident about a suboptimal prediction. In this work, we enhance beam search by performing minimum Bayes-risk (MBR) decoding for some extra steps at a later stage. In our experiments, we found that the conventional MBR r...

Minimum Bayes Risk (MBR) speech recognizers have been shown to yield improvements over the conventional maximum a-posteriori probability (MAP) decoders in the context of Nbest list rescoring and A search over recognition lattices. Segmental MBR (SMBR) procedures have been developed to simplify implementation of MBR recognizers, by segmenting the N-best list or lattice, to reduce the size of the...

In this paper, we propose a broad framework for Memory-Based Reasoning (MBR). The framework allows us to clarify the technical merits and limitations of several current MBR methods and to design new variants. The proposed computational framework consists of three components: a speciication language to deene an adaptive notion of relevant context for a query; mechanisms for retrieving this conte...

In this paper, we propose a probabilistic framework for memory-based reasoning (MBR). The framework allows us to clarify the technical merits and limitations of several recently published MBR methods and to design new variants. The proposed computational framework consists of three components: a specification language to define an adaptive notion of relevant context for a query; mechanisms for ...

This study considers movement-based registration (MBR) with implicit registration, an improved version of original MBR. In order to obtain more accurate registration cost of the MBR, we consider real network architecture in which a mobile network is composed of visitor location registers (VLR), a group of cells. We show that, when the network architecture is considered, the location registratio...

Minimum Bayes-Risk (MBR) speech recognizers have been shown to give improvements over the conventional maximum a-posteriori probability (MAP) decoders through N-best list rescoring and A search over word lattices. Segmental MBR (SMBR) decoders simplify the implementation of MBR recognizers by segmenting the N-best lists or lattices over which the recognition is performed. We present a lattice c...

The computation of the Minimum Bayes Risk (MBR) decoding rule for word lattices needs approximations. We investigate a class of approximations where the Levenshtein alignment is approximated under the condition that competing lattice arcs overlap in time. The approximations have their origins in MBR decoding and in discriminative training. We develop modified versions and propose a new, concept...

OBJECTIVE To determine which states have laws that require the review of mandated health insurance benefits and describe the various approaches states take in reviewing mandated benefits, as stated in the mandated benefit review (MBR) laws. DATA SOURCES We queried online databases of the individual state statutes and reviewed the state statutes and state legislative agendas for all 50 states ...

The influence of sludge properties in SMEBR and conventional MBR pilot systems on membrane fouling was investigated. Generated data were analyzed using statistical analysis Pearson's product momentum correlation coefficient (r(p)). Analysis showed that TMP had strong direct (r(p)=0.9182) and inverse (r(p)=-0.9205) correlations to mean particle size diameter in MBR and SMEBR, respectively. TMP i...