New framework enhances adaptive experimental design for complex trials

Researchers have developed a new framework for designing longitudinal adaptive experiments, particularly useful in clinical trials and digital experiments. This framework aims to improve estimation efficiency by dynamically updating treatment randomization probabilities as data accumulates. The proposed method includes an efficiency-oriented design criterion that can optimize for single or multiple estimands and a longitudinal targeted maximum likelihood estimator (ADL-LTMLE) for accurate estimation without parametric assumptions. Simulation studies indicate substantial variance reductions compared to non-adaptive designs, achieving performance close to the oracle design. AI

IMPACT Enhances experimental design and estimation efficiency in data-driven research, potentially improving AI model training and evaluation.

RANK_REASON Academic paper detailing a new statistical methodology. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv stat.ML →

• ADL-LTMLE
• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• DagsHub
• Gotit.pub
• Hugging Face
• Litmaps
• Longitudinal Adaptive Experimental Design for Learning Multiple Target Estimands with Semiparametric Efficient Inference
• ScienceCast
• scite Smart Citations

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →