Abstract. We propose a vector representation approach to contour estimation from noisy data. Images are modeled as random elds composed of a set of homogeneous regions contours (bo...
Whole-History Rating (WHR) is a new method to estimate the time-varying strengths of players involved in paired comparisons. Like many variations of the Elo rating system, the whol...
Inductive Logic Programming (ILP) involves the construction of first-order definite clause theories from examples and background knowledge. Unlike both traditional Machine Learnin...
Background: The estimation of demographic parameters from genetic data often requires the computation of likelihoods. However, the likelihood function is computationally intractab...
Daniel Wegmann, Christoph Leuenberger, Samuel Neue...
Dynamic Bayesian networks (DBNs) offer an elegant way to integrate various aspects of language in one model. Many existing algorithms developed for learning and inference in DBNs ...