Cost Sensitive Multi-Granulation Preference Relation Rough Set for Ordinal Decision System

The preference analysis is a class of important issues in multi-criteria ordinal decision making. The rough set is an effective approach to handle preference analysis. In order to solve the multi-criteria preference analysis problems, this paper improves the preference relation rough set model and expands it to multi-granulation cases. Cost is also an important issue in the field of decision analysis. Taking the cost into consideration, we also expand the model to the cost sensitive multi-granulation preference relation rough set. Some theorems are represented, and the granule structure selection based on approximation quality is investigated. The experimental results show that the multi-granulation preference rough set approach with the consideration of cost has a better performance in granule structure selection than that without cost consideration.