Abstract— Particle filters have been applied with great success to various state estimation problems in robotics. However, particle filters often require extensive parameter tw...
We address the problem of learning the mapping between words and their possible pronunciations in terms of sub-word units. Most previous approaches have involved generative modeli...
Spectral clustering and eigenvector-based methods have become increasingly popular in segmentation and recognition. Although the choice of the pairwise similarity metric (or affin...
Abstract a paradigm of modern Machine Learning (ML) which uses rewards and punishments to guide the learning process. One of the central ideas of RL is learning by “direct-online...
Background: Complex human diseases are often caused by multiple mutations, each of which contributes only a minor effect to the disease phenotype. To study the basis for these com...
Michael R. Mehan, Juan Nunez-Iglesias, Chao Dai, M...