Stochastic approach to rough set theory

Abstract

The presentation introduces the basic ideas and investigates the stochastic approach to rough set theory. The major aspects of the stochastic approach to rough set theory to be explored during the presentation are: the probabilistic view of the approximation space, the probabilistic approximations of sets, as expressed via variable precision and Bayesian rough set models, and probabilistic dependencies between sets and multi-valued attributes, as expressed by the absolute certainty gain and expected certainty gain measures, respectively. The measures allow for more comprehensive evaluation of rules computed from data and for computation of attribute reduct, core and significance factors in probabilistic decision tables. © Springer-Verlag Berlin Heidelberg 2006.