Detecting non-stationarity of a data-generating process is
challenging, particularly when the process is unknown and the only
information concerning the process status has to be extracted from a
set of observations.

This seminar presents a novel approach to change detection that
exploits the Intersection of Confidence Intervals (ICI) rule to
monitor the process evolution. The proposed change detection test can
also identify, after each detection, a set of observation generated
by the process in the novel status. These observations are used to
reconfigure the test, in order to promptly detect further
non-stationarities with respect to the current status.

Experimental results show that the proposed test outperforms
state-of-the-art solutions, both in terms of efficiency and
effectiveness, in particular when a reduced test configuration set is
available.