Information Geometry and Minimum Description Length Networks

Ke Sun, Jun Wang, Alexandros Kalousis, Stephan Marchand-Maillet
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:49-58, 2015.

Abstract

We study parametric unsupervised mixture learning. We measure the loss of intrinsic information from the observations to complex mixture models, and then to simple mixture models. We present a geometric picture, where all these representations are regarded as free points in the space of probability distributions. Based on minimum description length, we derive a simple geometric principle to learn all these models together. We present a new learning machine with theories, algorithms, and simulations.

Cite this Paper


BibTeX
@InProceedings{pmlr-v37-suna15, title = {Information Geometry and Minimum Description Length Networks}, author = {Ke Sun and Jun Wang and Alexandros Kalousis and Stephan Marchand-Maillet}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {49--58}, year = {2015}, editor = {Francis Bach and David Blei}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/suna15.pdf}, url = { http://proceedings.mlr.press/v37/suna15.html }, abstract = {We study parametric unsupervised mixture learning. We measure the loss of intrinsic information from the observations to complex mixture models, and then to simple mixture models. We present a geometric picture, where all these representations are regarded as free points in the space of probability distributions. Based on minimum description length, we derive a simple geometric principle to learn all these models together. We present a new learning machine with theories, algorithms, and simulations.} }
Endnote
%0 Conference Paper %T Information Geometry and Minimum Description Length Networks %A Ke Sun %A Jun Wang %A Alexandros Kalousis %A Stephan Marchand-Maillet %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-suna15 %I PMLR %P 49--58 %U http://proceedings.mlr.press/v37/suna15.html %V 37 %X We study parametric unsupervised mixture learning. We measure the loss of intrinsic information from the observations to complex mixture models, and then to simple mixture models. We present a geometric picture, where all these representations are regarded as free points in the space of probability distributions. Based on minimum description length, we derive a simple geometric principle to learn all these models together. We present a new learning machine with theories, algorithms, and simulations.
RIS
TY - CPAPER TI - Information Geometry and Minimum Description Length Networks AU - Ke Sun AU - Jun Wang AU - Alexandros Kalousis AU - Stephan Marchand-Maillet BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-suna15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 49 EP - 58 L1 - http://proceedings.mlr.press/v37/suna15.pdf UR - http://proceedings.mlr.press/v37/suna15.html AB - We study parametric unsupervised mixture learning. We measure the loss of intrinsic information from the observations to complex mixture models, and then to simple mixture models. We present a geometric picture, where all these representations are regarded as free points in the space of probability distributions. Based on minimum description length, we derive a simple geometric principle to learn all these models together. We present a new learning machine with theories, algorithms, and simulations. ER -
APA
Sun, K., Wang, J., Kalousis, A. & Marchand-Maillet, S.. (2015). Information Geometry and Minimum Description Length Networks. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:49-58 Available from http://proceedings.mlr.press/v37/suna15.html .

Related Material