AI Paper Atlas

Research map / Reinforcement learning / MDPs and planning

Hardness in Markov Decision Processes: Theory and Practice

Michelangelo Conserva, Paulo Rauber

NeurIPS 2022 · Reinforcement learning › MDPs and planning · approach: Average-reward and regularized MDPs · 0 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
Increase p_randMiniGridEmpty · Episodic MiniGridEmpty, scenario 1diameterHurt
Increase p_randMiniGridEmpty · Episodic MiniGridEmpty, scenario 1environmental value normImproved
Increase p_lazyMiniGridEmpty · Episodic MiniGridEmpty, scenario 2diameterHurt
Increase p_lazyMiniGridEmpty · Episodic MiniGridEmpty, scenario 2environmental value normImproved
Increase the number of statesMiniGridEmpty · Episodic MiniGridEmpty, scenarios 3 and 4cumulative regretHurt
Increase the number of statesMiniGridEmpty · Episodic MiniGridEmpty, scenarios 3 and 4environmental value normMixed
Increase p_rand or p_lazyMiniGridEmpty · Episodic MiniGridEmpty, scenarios 1 and 2cumulative regretMixed
Use the episodic setting rather than the communicating settingColosseum agents · Colosseum benchmarknormalized cumulative regretImproved
Use PSRL instead of Q-learningTabular agents · Episodic Colosseum benchmarknormalized cumulative regret: .81±.24 -> .30±.33; .83±.23 -> .35±.30Improved
Use UCRL2 instead of Q-learningTabular agents · Continuous ergodic Colosseum benchmarknormalized cumulative regret: .85±.18 -> .17±.26Improved
Use PSRL instead of Q-learningTabular agents · Continuous communicating Colosseum benchmarknormalized cumulative regret: .38±.37 -> .69±.38Hurt
Use non-tabular DQN rather than the other non-tabular agentsbsuite agents · Continuous ergodic Colosseum environmentscumulative regretMixed

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