Change-point identifications by means of genetic algorithms: a case study