AI Paper Atlas

Research map / Bandits and online learning / Contextual bandits and Thompson sampling

On Multi-Armed Bandit Designs for Dose-Finding Trials

Maryam Aziz, Emilie Kaufmann, Marie-Karelle Riviere

JMLR 2021 · Bandits and online learning › Contextual bandits and Thompson sampling · approach: Thompson sampling analysis · 6 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
add Thompson Sampling with logistic model in place of CRMTS · MTD identification, 9 scenariosrecommendation of the MTD: 3 out of 9Mixed
add epsilon regularization to Thompson Sampling (ablation)TS(epsilon) · MTD identification, 9 scenariosrecommendation of the MTD vs CRM: 5 out of 9Mixed
add admissible set to Thompson Sampling (ablation)TS A · MTD identification, 9 scenariosrecommendation of the MTD vs CRM: 5 out of 9Mixed
exploit increasing-toxicity structure via logistic model instead of independent uniform priorsTS · MTD identification, n=36recommendation of the MTD: 37.6 -> 78.9 (Scenario 1)Improved
add admissible set A or epsilon regularity to Thompson Sampling (ablation)TS A and TS(epsilon) · MTD identificationallocation to high-toxicity doses: consistently reduces the allocation percentage of higher toxicity dosesImproved
add admissible set to Thompson Sampling (ablation)TS A · MTD identification Sc. 3allocation % to MTD (dose 6): 43.2 -> 32.1Hurt
use 3+3 design instead of model-based designs3+3 · MTD identificationdose recommendation: 35.2 vs 77.2 (Scenario 1, dose 1)Hurt
vary epsilon in TS(epsilon) over 0.02, 0.05, 0.1 (ablation)TS(epsilon) · MTD identificationrecommendation performance: 0.02, 0.05, 0.1 comparableNo clear effect
vary c1 in TS A over 0.6, 0.8, 0.9 (ablation)TS A · MTD identificationrecommendation performance: 0.6, 0.8, 0.9 comparableNo clear effect
replace MTA-RA with TS ATS A · MED identification, scenarios 1-4 and 12-13optimal decision rate vs MTA-RA: 5 out of 6Improved
replace MTA-RA with TSTS · MED identification, scenarios 1-4 and 12-13optimal decision rate vs MTA-RA: 4 out of 6Improved
add admissible set to Thompson Sampling (ablation)TS A · MED identificationallocation to doses with toxicity above theta: always smaller for TS A than for MTA-RAImproved

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