AI Paper Atlas

Research map / Reinforcement learning / MDPs and planning

Approximate Newton Methods for Policy Search in Markov Decision Processes

Thomas Furmston, Guy Lever, David Barber

JMLR 2016 · Reinforcement learning › MDPs and planning · approach: Robust MDP policy gradient · 21 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 second Gauss-Newton method instead of EMpolicy search · Ball-in-a-cup robot armsuccessful runs out of 50: 36 -> 45Improved
use second Gauss-Newton method instead of natural gradient ascentpolicy search · 3-link rigid manipulatornormalized total expected reward: natural gradient ascent obtained around 15% of the performanceImproved
use second Gauss-Newton method instead of EMpolicy search · 3-link rigid manipulatornormalized total expected reward: EM obtained 40% of the performance; 50 seconds vs 300 secondsImproved
tune step size in Gauss-Newton rather than constant step 1.0 as in EM (ablation)second Gauss-Newton method · 3-link rigid manipulatornormalized total expected rewardImproved
use gradient ascentpolicy search · 3-link rigid manipulatornormalized total expected rewardHurt
use full Gauss-Newton instead of diagonal Gauss-Newton (ablation)Gauss-Newton policy search · Tetris 10x10 boardcompleted linesImproved
use diagonal Gauss-Newton instead of natural gradient ascentpolicy search · Tetris 10x10 boardcompleted linesNo clear effect
use second Gauss-Newton method instead of EMpolicy search · Ball-in-a-cup robot armtotal expected rewardImproved
use first Gauss-Newton method, gradient ascent or natural gradient ascentpolicy search · Ball-in-a-cup robot armtask successHurt

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