84 search results - page 17 / 17 » Negative Training Data Can be Harmful to Text Classification |
BMCBI
2005
13 years 5 months ago
2005
Background: The multitude of motif detection algorithms developed to date have largely focused on the detection of patterns in primary sequence. Since sequence-dependent DNA struc...
WWW
2009
ACM
14 years 6 months ago
2009
ACM
Transferring knowledge from one domain to another is challenging due to a number of reasons. Since both conditional and marginal distribution of the training data and test data ar...
KDD
2009
ACM
14 years 6 months ago
2009
ACM
Topic models provide a powerful tool for analyzing large text collections by representing high dimensional data in a low dimensional subspace. Fitting a topic model given a set of...
BMCBI
2008
13 years 5 months ago
2008
Background: Although there are a large number of thesauri for the biomedical domain many of them lack coverage in terms and their variant forms. Automatic thesaurus construction b...