Multi-task classification with infinite local experts
Publication
, Journal Article
Wang, C; An, Q; Carin, L; Dunson, DB
Published in: ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings
September 23, 2009
We propose a multi-task learning (MTL) framework for nonlinear classification, based on an infinite set of local experts in feature space. The usage of local experts enables sharing at the expert-level, encouraging the borrowing of information even if tasks are similar only in subregions of feature space. A kernel stick-breaking process (KSBP) prior is imposed on the underlying distribution of class labels, so that the number of experts is inferred in the posterior and thus model selection issues are avoided. The MTL is implemented by imposing a Dirichlet process (DP) prior on a layer above the task- dependent KSBPs. ©2009 IEEE.
Duke Scholars
Published In
ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings
DOI
ISSN
1520-6149
Publication Date
September 23, 2009
Start / End Page
1569 / 1572
Citation
APA
Chicago
ICMJE
MLA
NLM
Wang, C., An, Q., Carin, L., & Dunson, D. B. (2009). Multi-task classification with infinite local experts. ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 1569–1572. https://doi.org/10.1109/ICASSP.2009.4959897
Wang, C., Q. An, L. Carin, and D. B. Dunson. “Multi-task classification with infinite local experts.” ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, September 23, 2009, 1569–72. https://doi.org/10.1109/ICASSP.2009.4959897.
Wang C, An Q, Carin L, Dunson DB. Multi-task classification with infinite local experts. ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings. 2009 Sep 23;1569–72.
Wang, C., et al. “Multi-task classification with infinite local experts.” ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, Sept. 2009, pp. 1569–72. Scopus, doi:10.1109/ICASSP.2009.4959897.
Wang C, An Q, Carin L, Dunson DB. Multi-task classification with infinite local experts. ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings. 2009 Sep 23;1569–1572.
Published In
ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings
DOI
ISSN
1520-6149
Publication Date
September 23, 2009
Start / End Page
1569 / 1572