activity
20152022
most citedA Brief Survey of Deep Reinforcement Learning

4.4k citations · 4.6k across the 14 of their papers we have counts for

collaborators
Showing cs.LGShow all

10 papers · 1 filter

cs.LG20221 cited

REMuS-GNN: A Rotation-Equivariant Model for Simulating Continuum Dynamics

Mario Lino, Stati Fotiadis, Anil A. Bharath +1

Numerical simulation is an essential tool in many areas of science and engineering, but its performance often limits application in practice or when used to explore large parameter…

cs.LG20213 cited

Diversity-based Trajectory and Goal Selection with Hindsight Experience Replay

Tianhong Dai, Hengyan Liu, Kai Arulkumaran +2

Hindsight experience replay (HER) is a goal relabelling technique typically used with off-policy deep reinforcement learning algorithms to solve goal-oriented tasks; it is well sui…

cs.LG20213 cited

Simulating Continuum Mechanics with Multi-Scale Graph Neural Networks

Mario Lino, Chris Cantwell, Anil A. Bharath +1

Continuum mechanics simulators, numerically solving one or more partial differential equations, are essential tools in many areas of science and engineering, but their performance…

cs.LG20209 cited

Simulating Surface Wave Dynamics with Convolutional Networks

Mario Lino, Chris Cantwell, Stathi Fotiadis +2

We investigate the performance of fully convolutional networks to simulate the motion and interaction of surface waves in open and closed complex geometries. We focus on a U-Net ar…

cs.LG2020

Comparing recurrent and convolutional neural networks for predicting wave propagation

Stathi Fotiadis, Eduardo Pignatelli, Mario Lino Valencia +3

Dynamical systems can be modelled by partial differential equations and numerical computations are used everywhere in science and engineering. In this work, we investigate the perf…

cs.LG20192 cited

Sample-Efficient Reinforcement Learning with Maximum Entropy Mellowmax Episodic Control

Marta Sarrico, Kai Arulkumaran, Andrea Agostinelli +2

Deep networks have enabled reinforcement learning to scale to more complex and challenging domains, but these methods typically require large quantities of training data. An altern…