R
Ruslan Salakhutdinov
Researcher at Carnegie Mellon University
Publications - 457
Citations - 142495
Ruslan Salakhutdinov is an academic researcher from Carnegie Mellon University. The author has contributed to research in topics: Computer science & Artificial neural network. The author has an hindex of 107, co-authored 410 publications receiving 115921 citations. Previous affiliations of Ruslan Salakhutdinov include Carnegie Learning & University of Toronto.
Papers
More filters
Posted Content
A Generic Approach for Escaping Saddle points
Sashank J. Reddi,Manzil Zaheer,Suvrit Sra,Barnabás Póczos,Francis Bach,Ruslan Salakhutdinov,Alexander J. Smola +6 more
TL;DR: In this article, the authors propose a framework that alternates between a first-order and a second-order subroutine, using the latter only close to saddle points, and yields convergence results competitive to the state-of-the-art.
Posted Content
Learning Representations from Imperfect Time Series Data via Tensor Rank Regularization
Paul Pu Liang,Zhun Liu,Yao-Hung Hubert Tsai,Qibin Zhao,Ruslan Salakhutdinov,Louis-Philippe Morency +5 more
TL;DR: A regularization method based on tensor rank minimization is presented based on the observation that high-dimensional multimodal time series data often exhibit correlations across time and modalities which leads to low-rank tensor representations.
Proceedings Article
Capsules with Inverted Dot-Product Attention Routing
TL;DR: A new routing algorithm for capsule networks is introduced, in which a child capsule is routed to a parent based only on agreement between the parent's state and the child's vote, which improves performance on benchmark datasets and performs at-par with a powerful CNN with 4x fewer parameters.
Posted Content
How Many Samples are Needed to Learn a Convolutional Neural Network
TL;DR: It is shown that for learning an $m-dimensional convolutional filter with linear activation acting on a $d$-dimensional input, the sample complexity of achieving population prediction error of $\epsilon$ is $\widetilde{O} (m/\Epsilon^2)$, whereas its FNN counterpart needs at least $\Omega(d/\epsil on)$ samples.