B1882
Title: Inference in high-dimensional online changepoint detection
Authors: Yudong Chen - London School of Economics and Political Science (United Kingdom) [presenting]
Tengyao Wang - London School of Economics (United Kingdom)
Richard Samworth - University of Cambridge (United Kingdom)
Abstract: Two new inferential challenges are introduced and studied, which are associated with the sequential detection of change in a high-dimensional mean vector. First, a confidence interval is sought for the changepoint, and second, the set of indices of coordinates is estimated in which the mean changes. An online algorithm is proposed that produces an interval with guaranteed nominal coverage, and whose length is, with high probability, of the same order as the average detection delay, up to a logarithmic factor. The corresponding support estimate enjoys control of both false negatives and false positives. Simulations confirm the effectiveness of the methodology, and its applicability is also illustrated in the US excess deaths data from 2017-2020.
