To aggregate or to eliminate? Optimal model simplification for improved process performance prediction

Arik Senderovich, Alexander Shleyfman, Matthias Weidlich, Avigdor Gal, Avishai Mandelbaum

Abstract: Highlights•A technique for performance-driven model reduction of GSPNs is proposed.•The technique relies on foldings that aggregate or eliminate performance information.•Foldings preserve model stability and have a bound for the introduced performance estimation error.•Given a budget for the estimation error, an optimal sequence of foldings can be found.