We provide a reference implementation of our approach in the file change_point_identification.py along with an example to initialize the synthesizer and run the algorithm.
The file contains two classes. The class Synthesizer is used to synthesize and retrieve performance data for a controlled experiment setup. The class ChangePointLearner implements the learning algorithm.
The file requirements.txt contains a list of dependencies. These can be installed, for instance with pip via pip3 install -r requirements.txt.