114 search results - page 15 / 23 » jair 2006 |
JAIR
2006
13 years 9 months ago
2006
Loopy and generalized belief propagation are popular algorithms for approximate inference in Markov random fields and Bayesian networks. Fixed points of these algorithms have been...
JAIR
2006
13 years 9 months ago
2006
A -model is a satisfying assignment of a Boolean formula for which any small alteration, such as a single bit flip, can be repaired by flips to some small number of other bits, yi...
JAIR
2006
13 years 9 months ago
2006
The most basic assumption used in statistical learning theory is that training data and test data are drawn from the same underlying distribution. Unfortunately, in many applicati...
JAIR
2006
13 years 9 months ago
2006
We characterize the search landscape of random instances of the job shop scheduling problem (JSP). Specifically, we investigate how the expected values of (1) backbone size, (2) d...
JAIR
2006
13 years 9 months ago
2006
In this paper we propose a data intensive approach for inferring sentence-internal temporal relations. Temporal inference is relevant for practical NLP applications which either e...