Deep Tempering
نویسندگان
چکیده
Restricted Boltzmann Machines (RBMs) are one of the fundamental building blocks of deep learning. Approximate maximum likelihood training of RBMs typically necessitates sampling from these models. In many training scenarios, computationally efficient Gibbs sampling procedures are crippled by poor mixing. In this work we propose a novel method of sampling from Boltzmann machines that demonstrates a computationally efficient way to promote mixing. Our approach leverages an under-appreciated property of deep generative models such as the Deep Belief Network (DBN), where Gibbs sampling from deeper levels of the latent variable hierarchy results in dramatically increased ergodicity. Our approach is thus to train an auxiliary latent hierarchical model, based on the DBN. When used in conjunction with parallel-tempering, the method is asymptotically guaranteed to simulate samples from the target RBM. Experimental results confirm the effectiveness of this sampling strategy in the context of RBM training.
منابع مشابه
The effect of deep cryogenic treatment on mechanical properties of 80CrMo12 5 tool steel
Cryogenic treatment can be used as a supplemental treatment that is performed on some tool steels between quenching and tempering as an effective method for decreasing retained austenite and increasing wear resistance. In this research, the effect of deep cryogenic treatment (DCT) on dimensional stability and mechanical properties of 80CrMo12 5 tool steel was investigated. The martensitic trans...
متن کاملUsing deep sequencing to characterize the biophysical mechanism of a transcriptional regulatory sequence: SI Appendix
1. Supporting experimental procedures a. Plasmid library construction b. Media and growth rates c. Amplicon generation d. Processing of sequence reads e. Library substitution rates f. Post-sort loss of library diversity 2. Supporting theoretical methods a. Overview of mutual information b. Statistical inference using mutual information (justification of Eq. 1) c. Maximizing mutual information l...
متن کاملInvestigating the Effect of the Deep Cryogenic Heat Treatment on the Mechanical Properties and Corrosion Behavior of 1.2080 Tool Steel
Deep cryogenic heat treatment is assumed as a supplementary heat treatment performed on steels before the final tempering treatment to enhance the wear resistance and hardness of the steels. In this study, the effects of the deep cryogenic heat treatment on the wear behavior and corrosion resistance of the 1.2080 tool steel were studied using the wear testing machine and polarization and impeda...
متن کاملTraining Restricted Boltzmann Machines with Multi-tempering: Harnessing Parallelization
Restricted Boltzmann Machines (RBM’s) are unsupervised probabilistic neural networks that can be stacked to form Deep Belief Networks. Given the recent popularity of RBM’s and the increasing availability of parallel computing architectures, it becomes interesting to investigate learning algorithms for RBM’s that benefit from parallel computations. In this paper, we look at two extensions of the...
متن کاملParallel Tempering for Training of Restricted Boltzmann Machines
Alternating Gibbs sampling between visible and latent units is the most common scheme used for sampling from Restricted Boltzmann Machines (RBM), a crucial component in deep architectures such as Deep Belief Networks (DBN). However, we find that it often does a very poor job of rendering the diversity of modes captured by the trained model. We suspect that this property hinders RBM training met...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید
ثبت ناماگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید
ورودعنوان ژورنال:
- CoRR
دوره abs/1410.0123 شماره
صفحات -
تاریخ انتشار 2014