R
Robert Gens
Researcher at University of Washington
Publications - 8
Citations - 869
Robert Gens is an academic researcher from University of Washington. The author has contributed to research in topics: Bayesian network & Graphical model. The author has an hindex of 7, co-authored 8 publications receiving 798 citations.
Papers
More filters
Proceedings Article
Deep Symmetry Networks
Robert Gens,Pedro Domingos +1 more
TL;DR: Deep symmetry networks (symnets), a generalization of convnets that forms feature maps over arbitrary symmetry groups that uses kernel-based interpolation to tractably tie parameters and pool over symmetry spaces of any dimension are introduced.
Proceedings Article
Learning the Structure of Sum-Product Networks
Robert Gens,Domingos Pedro +1 more
TL;DR: This work proposes the first algorithm for learning the structure of SPNs that takes full advantage of their expressiveness, and shows that the learned SPNs are typically comparable to graphical models in likelihood but superior in inference speed and accuracy.
Proceedings Article
Discriminative Learning of Sum-Product Networks
Robert Gens,Pedro Domingos +1 more
TL;DR: This paper presents the first discriminative training algorithms for SPNs, combining the high accuracy of the former with the representational power and tractability of the latter, and proposes an efficient backpropagation-style algorithm for computing the gradient of the conditional log likelihood.
Journal ArticleDOI
On the Latent Variable Interpretation in Sum-Product Networks
TL;DR: It is shown that the Viterbi-style algorithm for MPE proposed in literature was never proven to be correct, and is found to be a sound derivation of the EM algorithm for SPNs when applied to augmented SPNs.
Learning Selective Sum-Product Networks
TL;DR: This work considers the selectivity constraint on the structure of sum-product networks, which allows each sum node to have at most one child with non-zero output for each possible input, to find globally optimal maximum likelihood parameters in closed form.