Experimental design for causal inference through an optimization lens
J Zhao - … in Operations Research: Smarter Decisions for a …, 2024 - pubsonline.informs.org
The study of experimental design offers tremendous benefits for answering causal questions
across a wide range of applications, including agricultural experiments, clinical trials …
across a wide range of applications, including agricultural experiments, clinical trials …
Beyond Neyman–Pearson: E-values enable hypothesis testing with a data-driven alpha
PD Grünwald - Proceedings of the National Academy of Sciences, 2024 - pnas.org
A standard practice in statistical hypothesis testing is to mention the P-value alongside the
accept/reject decision. We show the advantages of mentioning an e-value instead. With P …
accept/reject decision. We show the advantages of mentioning an e-value instead. With P …
Anytime-Valid Continuous-Time Confidence Processes for Inhomogeneous Poisson Processes
Motivated by monitoring the arrival of incoming adverse events such as customer support
calls or crash reports from users exposed to an experimental product change, we consider …
calls or crash reports from users exposed to an experimental product change, we consider …
YEAST: Yet Another Sequential Test
A Kurennoy, M Dodin, T Gurbanov… - arXiv preprint arXiv …, 2024 - arxiv.org
Large-scale randomised experiments have become a standard tool for developing products
and improving user experience. To reduce losses from shipping harmful changes …
and improving user experience. To reduce losses from shipping harmful changes …
Basemath's Test: Group Sequential Testing Without Recurrent Numerical Integration
A Weiss - Available at SSRN 4599695, 2024 - papers.ssrn.com
Basemath's Test is a novel (Group) Sequential Testing approach. This method simplifies
statistical analysis by eliminating the need for recurrent numerical integration. The presented …
statistical analysis by eliminating the need for recurrent numerical integration. The presented …