Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Harm van Seijen

Shortest-Path Constrained Reinforcement Learning for Sparse Reward Tasks

Jul 13, 2021
Sungryull Sohn, Sungtae Lee, Jongwook Choi, Harm van Seijen, Mehdi Fatemi, Honglak Lee

* In proceedings of ICML 2021 

  Access Paper or Ask Questions

A Deeper Look at Discounting Mismatch in Actor-Critic Algorithms

Oct 02, 2020
Shangtong Zhang, Romain Laroche, Harm van Seijen, Shimon Whiteson, Remi Tachet des Combes

  Access Paper or Ask Questions

The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning

Jul 07, 2020
Harm van Seijen, Hadi Nekoei, Evan Racah, Sarath Chandar

  Access Paper or Ask Questions

Using a Logarithmic Mapping to Enable Lower Discount Factors in Reinforcement Learning

Jun 03, 2019
Harm van Seijen, Mehdi Fatemi, Arash Tavakoli

  Access Paper or Ask Questions

Learning Invariances for Policy Generalization

Sep 07, 2018
Remi Tachet des Combes, Philip Bachman, Harm van Seijen

* 7 pages, 1 figure 

  Access Paper or Ask Questions

Hybrid Reward Architecture for Reinforcement Learning

Nov 28, 2017
Harm van Seijen, Mehdi Fatemi, Joshua Romoff, Romain Laroche, Tavian Barnes, Jeffrey Tsang

  Access Paper or Ask Questions

Multi-Advisor Reinforcement Learning

Nov 14, 2017
Romain Laroche, Mehdi Fatemi, Joshua Romoff, Harm van Seijen

* Submitted at ICLR2018 

  Access Paper or Ask Questions

Separation of Concerns in Reinforcement Learning

Mar 28, 2017
Harm van Seijen, Mehdi Fatemi, Joshua Romoff, Romain Laroche

  Access Paper or Ask Questions

True Online Temporal-Difference Learning

Sep 08, 2016
Harm van Seijen, A. Rupam Mahmood, Patrick M. Pilarski, Marlos C. Machado, Richard S. Sutton

* Journal of Machine Learning Research (JMLR), 17(145):1-40, 2016 
* This is the published JMLR version. It is a much improved version. The main changes are: 1) re-structuring of the article; 2) additional analysis on the forward view; 3) empirical comparison of traditional and new forward view; 4) added discussion of other true online papers; 5) updated discussion for non-linear function approximation 

  Access Paper or Ask Questions

Effective Multi-step Temporal-Difference Learning for Non-Linear Function Approximation

Aug 18, 2016
Harm van Seijen

  Access Paper or Ask Questions

An Empirical Evaluation of True Online TD(位)

Jul 01, 2015
Harm van Seijen, A. Rupam Mahmood, Patrick M. Pilarski, Richard S. Sutton

* European Workshop on Reinforcement Learning (EWRL) 2015 

  Access Paper or Ask Questions

Planning by Prioritized Sweeping with Small Backups

Jan 10, 2013
Harm van Seijen, Richard S. Sutton

  Access Paper or Ask Questions