4670 search results - page 803 / 934 » Testing that distributions are close |
165
click to vote
IMAGING
2004
15 years 8 months ago
2004
Multispectral printer characterization requires an effective model to map printer input digital counts into reflectance spectra and vice versa. This paper presents a novel strateg...
188
click to vote
DICTA
2003
15 years 8 months ago
2003
Abstract. Mahalanobis distance has already proved its strength in human skin detection using a set of skin values. We present this work that uses automatic skin detection after an ...
156
click to vote
NIPS
2004
15 years 8 months ago
2004
We propose a novel a framework for deriving approximations for intractable probabilistic models. This framework is based on a free energy (negative log marginal likelihood) and ca...
173
click to vote
NIPS
2001
15 years 8 months ago
2001
Estimating insurance premia from data is a difficult regression problem for several reasons: the large number of variables, many of which are discrete, and the very peculiar shape...
198
click to vote
NIPS
2001
15 years 8 months ago
2001
We propose a new approach to reinforcement learning which combines least squares function approximation with policy iteration. Our method is model-free and completely off policy. ...