Return-to-Baseline Testing via Empirically Calibrated e-processes
Marta Regis, Paulo Serra · 2026
Read Return-to-Baseline Testing via Empirically Calibrated e-processes by Marta Regis, Paulo Serra (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-01960.txt.
Description
Regis and Serra test when monitored data return to a pre-intervention baseline. High-frequency data are watched before and after an intervention. The goal is to spot when the source realigns. That moment marks a return to baseline. A good test must control errors at any stopping time.
Their procedure is sequential and distribution-free. No null model needs to be specified upfront. A discrepancy measure builds a non-negative supermartingale. It is then calibrated on the baseline data itself. Calibration is thus specific to each subject studied. Finite-sample bounds cover the calibration error. Simulations and a clinical case study show the method. Tuning choices and compute costs are also discussed.
Catalog record
| CREATOR | Marta Regis, Paulo Serra |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.01960