Conference Abstracts October 11-14, 2009 t provides the abstracts and the author(s)’ name(s) for the technical program. The actual presenter(s) may vary. Updated September 20, 20...
Multiple-instance Learning (MIL) is a new paradigm
of supervised learning that deals with the classification of
bags. Each bag is presented as a collection of instances
from whi...
Zhouyu Fu (Australian National University), Antoni...
Deformable model fitting has been actively pursued in the computer vision
community for over a decade. As a result, numerous approaches have
been proposed with varying degrees of...
We present a general Bayesian framework for hyperparameter tuning in L2-regularized supervised learning models. Paradoxically, our algorithm works by first analytically integratin...
In Web-based services of dynamic content (such as news articles), recommender systems face the difficulty of timely identifying new items of high-quality and providing recommendat...