… sigmoid belief networks, based on data augmentation. From the perspective of learning, we
desire … on the model parameters {W(l)} and {c(l)}, and distributions on the data-dependent {h …

… of normals to derive data augmentation strategies for deep learning. This allows variants of
… deep learning models. To demonstrate our methodology, we develop data augmentation …

… In Section 3.2 we demonstrate that augmenting Q with Pólya-Gamma (PG) random variables
leads to an analytic updating scheme that vastly improves the convergence speed of the M-…

… true of augmented EEG signals. In other words, correctly labeling augmented datasets can
… The earliest paper we could find that specifically used the name “data augmentation” was in …

… conditional generative models are developed to simultaneously learn the temporal depen…
the Temporal Sigmoid Belief Network (TSBN), a sequential stack of Sigmoid Belief Networks (…

… data augmentation to solve large-scale data repositories that are not publicly available. In
this study, deep learning … performance tuning of large-scale data repositories. We propose a …

… (2015b), using the Polya-Gamma data augmentation technique developed for logistic
models (Polson et al., 2012). In this paper, we will develop data augmentation technique unique …

… for dictionary learning, … Data augmentation and Kronecker methods allow for efficient
Markov chain Monte Carlo sampling. We further extend the model with Sigmoid Belief Networks (…

… three phases: Data augmentation to learn robust features, a pre-… is not factorizable in each
training case. Integrating the … makes learning layer-wise difficult in a sigmoid belief network.…

… [16] proposed Deep Belief Networks (DBNs) with a new unsupervised … The common data
augmentation methods include mirroring, scaling, and rotation. Aggressive data augmentation …