AI Paper Atlas

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

Deep Recurrent Optimal Stopping

NIRANJAN DAMERA VENKATA, Chiranjib Bhattacharyya

NeurIPS 2023 · Reinforcement learning › Policy gradient and actor-critic · approach: Alternative policy update rules · 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
Fit the stopping policy directly (OSPG) instead of fitted Q-iteration with a DNNDNN · Bermudan max-call d=20 p0=100average return: 50.92 -> 51.02Improved
Use a recurrent policy instead of feed-forward on Markovian dataOSPG · Bermudan max-call d=100 p0=100average return: 82.96 -> 82.74No clear effect
Use RRLSM with fixed random RNN features instead of OSPGRNN · Bermudan max-call d=20 p0=100average return: 50.99 -> 49.68Hurt
Use recurrent OSPG instead of DNN-OSPG on non-Markovian fBmOSPG · Fractional Brownian motion h=0.05average return: 1.14 -> 1.28Improved
Use OSPG instead of FQI with an RNN on fBmRNN · Fractional Brownian motion h=0.80average return: 0.18 -> 0.27Improved
Extend DOS state with process history (DOS-ES)DOS · Fractional Brownian motion h=0.05average return: 0.70 -> 1.18Improved
Directly fit policy (DNN-OSPG) vs value-recursion methods DOS and DNN-FQIDNN · Fractional Brownian motion h=0.05average return: 0.87 -> 1.14Improved
Use DNN-OSPG instead of DNN-FQI on American geometric-call optionDNN · American geometric-call d=7 s0=110average return: 15.0487 -> 15.9699Improved
Use OSPG versus Longstaff-Schwartz at higher asset countDNN-OSPG · American geometric-call d=13 s0=90average return: 5.5962 -> 5.7436Mixed
Use RNN-OSPG instead of RRLSM for early stopping classifierRNN · UCR CBF early classificationaverage cost: 0.192 -> 0.186Improved
Use RNN-OSPG instead of RNN-FQI for early stoppingRNN · UCR CBF early classificationaverage cost: 0.308 -> 0.186Improved
Share parameters across time-steps (DNN-OSPG) vs DOS per-step networksDNN · Bermudan max-call d=20model size (params): 9,225 -> 1,047Improved

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