Abstract
Stochastic automata networks and generalized stochastic Petri nets are the main formalisms used to model complex Markov systems in a structured "Kronecker" approach. We compare them on a suite of examples using two tools, PEPS and SMART.