Current Trends in Rough Set Flow Graphs

Pawlak (2002, 2003) introduced rough set flow graphs (RSFGs) as a graphical framework for uncertainty management. RSFGs extend traditional rough set research (Pawlak, 1982, 1991) by organizing the rules obtained from decision tables as a directed acyclic graph (DAG). Each rule is associated with three coefficients, namely, strength, certainty, and coverage, that have been shown to satisfy Bayes’ theorem (Pawlak, 2002, 2003). Pawlak stated that RSFGs are a new perspective on Bayesian inference, and provided an algorithm to answer queries in a RSFG (Pawlak, 2003). We established in Butz, Yan, and Yang (2005) that a RSFG is, in fact, a special case of Bayesian network, and that RSFG can be carried out in polynomial time. In Butz et al. (2005), it was also shown that the traditional RSFG inference AbstrAct