papers

Publications (6)

cs.NE2019

A Dynamically Controlled Recurrent Neural Network for Modeling Dynamical Systems

Yiwei Fu, Samer Saab, Asok Ray +1

This work proposes a novel neural network architecture, called the Dynamically Controlled Recurrent Neural Network (DCRNN), specifically designed to model dynamical systems that ar…

cs.NE2019

State Space Representations of Deep Neural Networks

Michael Hauser, Sean Gunn, Samer Saab +1

This paper deals with neural networks as dynamical systems governed by differential or difference equations. It shows that the introduction of skip connections into network archite…

astro-ph.CO2009

A New Era in Extragalactic Background Light Measurements: The Cosmic History of Accretion, Nucleosynthesis and Reionization

Asantha Cooray, Alexandre Amblard, Charles Beichman +52

(Brief Summary) What is the total radiative content of the Universe since the epoch of recombination? The extragalactic background light (EBL) spectrum captures the redshifted ener…

cs.LG2019

Training products of expert capsules with mixing by dynamic routing

Michael Hauser

This study develops an unsupervised learning algorithm for products of expert capsules with dynamic routing. Analogous to binary-valued neurons in Restricted Boltzmann Machines, th…

cs.NE2019

On Residual Networks Learning a Perturbation from Identity

Michael Hauser

The purpose of this work is to test and study the hypothesis that residual networks are learning a perturbation from identity. Residual networks are enormously important deep learn…

cs.NE2019

Training capsules as a routing-weighted product of expert neurons

Michael Hauser

Capsules are the multidimensional analogue to scalar neurons in neural networks, and because they are multidimensional, much more complex routing schemes can be used to pass inform…