Tweaking the DiPasquale–Wheaton–Colwell model