论文信息 - Joint Training of Deep Boltzmann Machines - 字舞流文

Joint Training of Deep Boltzmann Machines

We introduce a new method for training deep Boltzmann machines jointly. Prior methods require an initial learning pass that trains the deep Boltzmann machine greedily, one layer at a time, or do not perform well on classifi- cation tasks.

Yoshua Bengio | Aaron C. Courville | Ian J. Goodfellow | Yoshua Bengio

[1] Nitish Srivastava,et al. Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[2] Veselin Stoyanov,et al. Empirical Risk Minimization of Graphical Model Parameters Given Approximate Inference, Decoding, and Model Structure , 2011, AISTATS.

[3] Tijmen Tieleman,et al. Training restricted Boltzmann machines using approximations to the likelihood gradient , 2008, ICML '08.

[4] L. Younes. On the convergence of markovian stochastic algorithms with rapidly decreasing ergodicity rates , 1999 .

[5] Geoffrey E. Hinton,et al. Deep Boltzmann Machines , 2009, AISTATS.

[6] Yann Ollivier,et al. Layer-wise learning of deep generative models , 2012, ArXiv.