Reverse prediction

Abstract

Rough set reverse method enables prediction of the best values for condition attributes given values for the decision attributes. Reverse prediction is required for many problems that do not lend themselves to being solved by the traditional rough sets forward prediction. The RS1 algorithm has been rewritten using better notation and style and generalized to provide reverse prediction. Rough Set Reverse Prediction Algorithm was implemented and evaluated on its ability to make inferences on large data sets in a dynamic problem domain. © Springer-Verlag Berlin Heidelberg 2005.