AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Fixed-Horizon Temporal Difference Methods for Stable Reinforcement Learning

Kristopher De Asis, Alan Chan, Silviu Pitis, Richard Sutton, Daniel Graves

AAAI 2020 · Reinforcement learning › Policy gradient and actor-critic · approach: Temporal-difference and discounting theory · 23 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use one-step fixed-horizon TD instead of conventional TDBaird's counterexample · Baird's counterexamplestabilityImproved
Use a final horizon of 16 rather than 8 (ablation)FHQ-learning · Slippery mazemean episode lengthImproved
Use a shorter horizon rather than the full episodic return (ablation)FHQ-learning and Q-learning · Slippery mazemean episode lengthImproved
Use fixed horizons instead of conventional discount ratesFHQ-learning versus Q-learning · Slippery mazemean episode lengthMixed
Use Deep FHQ-learning rather than DQNLunarLander-v2 · LunarLander-v2returnImproved
Set the discount factor to 1 rather than 0.99 (ablation)DQN · LunarLander-v2training stabilityHurt
Set the discount factor to 1 rather than 0.99 (ablation)Deep FHQ-learning · LunarLander-v2performance sensitivity to discount factorNo clear effect
Add target networks (ablation)Deep FHQ-learning · LunarLander-v2learning speedHurt
Add target networks (ablation)DQN · LunarLander-v2area under the learning curveImproved
Increase hidden-layer width beyond the reported setting (ablation)Deep FHQ-learning · LunarLander-v2performanceNo clear effect
Increase hidden-layer width beyond the reported setting (ablation)DQN · LunarLander-v2performanceHurt
Use a longer fixed horizon (ablation)FHTD · 19-state random walkroot-mean-squared errorMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy