Timezone: »

Training Restricted Boltzmann Machine via the Thouless-Anderson-Palmer free energy
Marylou Gabrie · Eric W Tramel · Florent Krzakala

Tue Dec 08 04:00 PM -- 08:59 PM (PST) @ 210 C #19

Restricted Boltzmann machines are undirected neural networks which have been shown tobe effective in many applications, including serving as initializations fortraining deep multi-layer neural networks. One of the main reasons for their success is theexistence of efficient and practical stochastic algorithms, such as contrastive divergence,for unsupervised training. We propose an alternative deterministic iterative procedure based on an improved mean field method from statistical physics known as the Thouless-Anderson-Palmer approach. We demonstrate that our algorithm provides performance equal to, and sometimes superior to, persistent contrastive divergence, while also providing a clear and easy to evaluate objective function. We believe that this strategycan be easily generalized to other models as well as to more accurate higher-order approximations, paving the way for systematic improvements in training Boltzmann machineswith hidden units.

Author Information

Marylou Gabrie (École Normale Supérieure, ICFP, Laboratoire de Physique Statistique )
Eric W Tramel (LPS, École Normale Supérieure)
Florent Krzakala (Ecole Normale Superieure Paris)

More from the Same Authors