Oxford University Press / Neural Information Processing Systems / Neural Networks / MIT Press / Nh Nv / GPU / Samsung / Curran Associates Inc. / /
Country
Tonga / France / Jordan / /
Facility
University of Toronto / /
IndustryTerm
online fashion / deep belief networks / optimization algorithms / optimization algorithm / recurrent networks / online learning / implicit matrix-vector products / sparse matrix-vector product / matrix-vector products / computing / natural gradient algorithm / linear systems / curvature matrix-vector products / representation learning algorithms / dynamical systems / large linear systems / approximately solving large linear systems / /
NaturalFeature
Lake et al. / /
Organization
University of Toronto Abstract Second-order / MIT / Oxford University / Inverse Fisher Matrix Roger B. Grosse Ruslan Salakhutdinov Department of Computer Science / University of Toronto / /
Person
James Martens / Y. Adaptive / Michael I. Graphical / Geoffrey Hinton / Yuri Burda / Ruslan / Murray / /
Scaling Up Natural Gradient by Sparsely Factorizing the Inverse Fisher Matrix Roger B. Grosse Ruslan Salakhutdinov Department of Computer Science, University of Toronto