[Research] Sequential Covariate Shift Detection

We propose a covariate shift detection algorithm in a sequential setting where data comes in a sequential order. This algorithm is based on classifier based two-sample tests and Clopper-Perason interval. We provide a False Postive Rate (FPR) bound and False Negative Rate (FNR) bound. This is published in ICML 2022, and the code is publicly released. Paper Github