Logic, Probability, and Rough Sets

摘要: The paper analyzes some properties of decision rules in the framework rough set theory and knowledge discovery systems. With every rule two conditional probabilities are associated called certatinty coverage factors, respectively. It is shown that these coefficients satisfy Bayes’ theorem. This relationship can be used as a new approach to Bayesian reasoning, without referring prior posterior probabilities, inherently with classical inference. Decision implications between theorem first was revealed by Lukasiewicz connection his multivaled logic.