Background: Sequencing of environmental DNA (often called metagenomics) has shown tremendous potential to uncover the vast number of unknown microbes that cannot be cultured and s...
Many unsupervised learning algorithms make use of kernels that rely on the Euclidean distance between two samples. However, the Euclidean distance is optimal for Gaussian distribut...
Karim T. Abou-Moustafa, Mohak Shah, Fernando De la...
In two-class score-based problems the combination of scores from an ensemble of experts is generally used to obtain distributions for positive and negative patterns that exhibit a ...
We conduct a series of Part-of-Speech (POS) Tagging experiments using Expectation Maximization (EM), Variational Bayes (VB) and Gibbs Sampling (GS) against the Chinese Penn Treeba...
We examine the learning-curve sampling method, an approach for applying machinelearning algorithms to large data sets. The approach is based on the observation that the computatio...