297 search results - page 44 / 60 » Learning from Ambiguously Labeled Examples |
AAAI
2012
12 years 1 months ago
2012
Relation extraction, the process of converting natural language text into structured knowledge, is increasingly important. Most successful techniques use supervised machine learni...
ICML
2010
IEEE
14 years 1 days ago
2010
IEEE
We present a novel approach for structure prediction that addresses the difficulty of obtaining labeled structures for training. We observe that structured output problems often h...
UAI
2003
14 years 10 days ago
2003
We formulate a principle for classification with the knowledge of the marginal distribution over the data points (unlabeled data). The principle is cast in terms of Tikhonov styl...
KDD
2005
ACM
14 years 11 months ago
2005
ACM
Discriminative sequential learning models like Conditional Random Fields (CRFs) have achieved significant success in several areas such as natural language processing, information...
ICML
2009
IEEE
14 years 11 months ago
2009
IEEE
We describe a new method for learning the conditional probability distribution of a binary-valued variable from labelled training examples. Our proposed Compositional Noisy-Logica...