Spoken dialog systems typically rely on recognition confidence scores to guard against potential misunderstandings. While confidence scores can provide an initial assessment for the reliability of the information obtained from the user, ideally systems should leverage information that is available in subsequent user responses to update and improve the accuracy of their beliefs. We present a mac...

This work is concerned with incrementally training statistical machine translation (SMT) models when new data becomes available. That, in contrast to re-training new models based on the entire accumulated data. Incremental training provides a way to perform faster, more frequent model updates, enabling keeping the SMT system up-to-date with the most recent data. Specifically, we address increme...

The Unifled Medical Language System (UMLS) is intended to support uniform access of machine-readable biomedical information resources. The foundation of the UMLS is a Metathesaurus, which will link terms in different biomedical nomenclatures. Because the resources and nomenclatures continue to evolve, the Metathesaurus must evolve with them. Thus, an important criterion for the design of the Me...

This paper describes a study of FE model verification. Before undertaking a model updating procedure, it is important to determine whether the initial model can be updated or not. Two methods for model verification a convergence check and a configuration check are proposed here to address the problems of discretisation and configuration errors in FE models, respectively. The differences between...

This paper models an agent in a three-period setting who does not update according to Bayes’Rule, and who is self-aware and anticipates her updating behavior when formulating plans. The agent is rational in the sense that her dynamic behavior is derived from a stable preference order on a domain of state-contingent menus of acts. A representation theorem generalizes (the dynamic version of) Ans...

The research presented in this thesis addresses the problem of updating an analytical model using a parametric Reference Basis approach. In this method, some parameters are assumed to be accurate (e.g. natural frequencies, mode shapes and mass matrix), while others are adjusted so that the eigenvalue equation is satisfied. Updating is done with the use of principal submatrices, and the method s...

SUMMARY New approaches are developed that use measured data to adjust the analytical mass and stiiness matrices of a system so that the agreement between the analytical modes of vibration and the modal survey is improved. By adding known masses to the structure of interest, measuring the modes of vibration of this mass-modiied system, and nally using this set of new data in conjunction with the...

Relevance reasoning in Bayesian networks can be used to improve eeciency of belief updating algorithms by identifying and pruning those parts of a network that are irrelevant for the computation. Relevance reasoning is based on the graphical property of d{separation and other simple and eecient techniques, the computational complexity of which is usually negligible when compared to the complexi...

We design and conduct an economic experiment to investigate the learning process of the agents under compound risk and under ambiguity. We gather data for subjects choosing between lotteries involving risky and ambiguous urns. Decisions are made in conjunction with a sequence of random draws with replacement, allowing us to estimate the beliefs of the agents at different moments in time. For ea...

We consider the problem of updating beliefs for binary random variables, when probability assessments are elicited for them based on information of varying quality. We propose the threshold model, a Bayesian updating procedure where only measures of location and correlation have to be speciied before any updating is possible. The main aspect of this model is the use of Jeerey's conditionalizati...