Reduct Equivalent Rule Induction Based On Rough Set Theory

Rough set theory is successfully used within data mining for prediction, classification (S. Abidi et al., 2001),(J.S. Deogun et al., 1995),(A. Kusiak, 2001),(P. Lingras, 2002),(P. Lingras et al., 2003), discovery of associations (J. Guan et al., 2003) and reduction of attributes (L. Mazlack et al., 2000), (M. Zhang and J.T. Yao, 2004). Having as motto "Let data speak for themselves", we can state that this approach is not invasive for the processed data set as it uses only the information of the data set without presuming the existence of different models in it (V. Raghavan and H. Sever, 1995). This article presents original contributions for the classifications of the objects of a database using the elements of rough set theory. Reduct equivalent rule induction (or RERT), is presented as a new classification method based on rough set theory. This article describes the mathematical essence of the reduct equivalent rule induction method, as well as the algorithm and the obtained experimental results.