Information Reduction Using Pseudo Consistency

Rough sets theory is a method to represent, analyze, and manipulate knowledge in an information table. To remove superfluous attributes without changing the original knowledge, reduction is a must in rough sets. The process of finding all reducts has been proven to be NP-hard. In order to find an efficient method to generate all reducts, this paper develops the pseudo consistency method. Simulations are generated and compared to the well known rough sets software – Rough Set Exploration System (RSES). The results show that the running time improvement can be up to thousands of times faster than the RSES. The larger size of attributes or instances is and the greater improvement can be achieved from the pseudo consistency method. In this paper, it does not only show the running time improvement but it also guarantees the final solutions are minimal without an extra verification process when they are found. This is not shown in most of the literatures. Applying the pseudo consistency method, this paper believes the rough sets theory can be efficiently extended to most of the real life applications.