Multi-view hidden markov perceptrons
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Standard
Lernen, Wissensentdeckung und Adaptivitat, LWA 2005. ed. / Mathias Bauer; Boris Brandherm; Johannes Fürnkranz; Gunter Grieser; Andreas Hotho; Andreas Jedlitschka; Alexander Kröner. Saarbrücken: Gesellschaft für Informatik e.V., 2005. p. 134-138.
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - CHAP
T1 - Multi-view hidden markov perceptrons
AU - Brefeld, Ulf
AU - Büscher, Christoph
AU - Scheffer, Tobias
PY - 2005
Y1 - 2005
N2 - Discriminative learning techniques for sequential data have proven to be more effective than generative models for named entity recognition, information extraction, and other tasks of discrimination. However, semi-supervised learning mechanisms that utilize inexpensive unlabeled sequences in addition to few labeled sequences – such as the Baum-Welch algorithm – are available only for generative models. The multi-view approach is based on the principle of maximizing the consensus among multiple independent hypotheses; we develop this principle into a semisupervised hidden Markov perceptron algorithm. Experiments reveal that the resulting procedure utilizes unlabeled data effectively and discriminates more accurately than its purely supervised counterparts.
AB - Discriminative learning techniques for sequential data have proven to be more effective than generative models for named entity recognition, information extraction, and other tasks of discrimination. However, semi-supervised learning mechanisms that utilize inexpensive unlabeled sequences in addition to few labeled sequences – such as the Baum-Welch algorithm – are available only for generative models. The multi-view approach is based on the principle of maximizing the consensus among multiple independent hypotheses; we develop this principle into a semisupervised hidden Markov perceptron algorithm. Experiments reveal that the resulting procedure utilizes unlabeled data effectively and discriminates more accurately than its purely supervised counterparts.
KW - Informatics
KW - Business informatics
UR - http://www.scopus.com/inward/record.url?scp=84874002288&partnerID=8YFLogxK
M3 - Article in conference proceedings
AN - SCOPUS:84874002288
SP - 134
EP - 138
BT - Lernen, Wissensentdeckung und Adaptivitat, LWA 2005
A2 - Bauer, Mathias
A2 - Brandherm, Boris
A2 - Fürnkranz, Johannes
A2 - Grieser, Gunter
A2 - Hotho, Andreas
A2 - Jedlitschka, Andreas
A2 - Kröner, Alexander
PB - Gesellschaft für Informatik e.V.
CY - Saarbrücken
T2 - Lernen, Wissensentdeckung und Adaptivitat, LWA 2005
Y2 - 10 October 2005 through 12 October 2005
ER -