Parametric Herding

Yutian Chen, Max Welling
Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, PMLR 9:97-104, 2010.

Abstract

A parametric version of herding is formulated. The nonlinear mapping between consecutive time slices is learned by a form of self-supervised training. The resulting dynamical system generates pseudo-samples that resemble the original data. We show how this parametric herding can be successfully used to compress a dataset consisting of binary digits. It is also verified that high compression rates translate into good prediction performance on unseen test data.

Cite this Paper


BibTeX
@InProceedings{pmlr-v9-chen10a, title = {Parametric Herding}, author = {Chen, Yutian and Welling, Max}, booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, pages = {97--104}, year = {2010}, editor = {Teh, Yee Whye and Titterington, Mike}, volume = {9}, series = {Proceedings of Machine Learning Research}, address = {Chia Laguna Resort, Sardinia, Italy}, month = {13--15 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v9/chen10a/chen10a.pdf}, url = { http://proceedings.mlr.press/v9/chen10a.html }, abstract = {A parametric version of herding is formulated. The nonlinear mapping between consecutive time slices is learned by a form of self-supervised training. The resulting dynamical system generates pseudo-samples that resemble the original data. We show how this parametric herding can be successfully used to compress a dataset consisting of binary digits. It is also verified that high compression rates translate into good prediction performance on unseen test data.} }
Endnote
%0 Conference Paper %T Parametric Herding %A Yutian Chen %A Max Welling %B Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2010 %E Yee Whye Teh %E Mike Titterington %F pmlr-v9-chen10a %I PMLR %P 97--104 %U http://proceedings.mlr.press/v9/chen10a.html %V 9 %X A parametric version of herding is formulated. The nonlinear mapping between consecutive time slices is learned by a form of self-supervised training. The resulting dynamical system generates pseudo-samples that resemble the original data. We show how this parametric herding can be successfully used to compress a dataset consisting of binary digits. It is also verified that high compression rates translate into good prediction performance on unseen test data.
RIS
TY - CPAPER TI - Parametric Herding AU - Yutian Chen AU - Max Welling BT - Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics DA - 2010/03/31 ED - Yee Whye Teh ED - Mike Titterington ID - pmlr-v9-chen10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 9 SP - 97 EP - 104 L1 - http://proceedings.mlr.press/v9/chen10a/chen10a.pdf UR - http://proceedings.mlr.press/v9/chen10a.html AB - A parametric version of herding is formulated. The nonlinear mapping between consecutive time slices is learned by a form of self-supervised training. The resulting dynamical system generates pseudo-samples that resemble the original data. We show how this parametric herding can be successfully used to compress a dataset consisting of binary digits. It is also verified that high compression rates translate into good prediction performance on unseen test data. ER -
APA
Chen, Y. & Welling, M.. (2010). Parametric Herding. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 9:97-104 Available from http://proceedings.mlr.press/v9/chen10a.html .

Related Material