Abstract

Z. Pawlak first proposed the rough set (RS) in 1982. For over forty years, scholars have developed a large number of RS models to solve various data problems. However, most RS models are designed based on inherent rules, and their mathematical structures are similar and complex. For this reason, the efficiency of RS methods in analyzing data has not been significantly improved. To address this issue, we propose some new rules to simplify traditional RS models. These simplified RS models, which are equivalent to traditional RS models, can mine data more quickly. In this paper, we take Pawlak RS as an example to compare the computational efficiency between the simplified Pawlak RS (SPRS) and the traditional RSs. Numerical experiments confirm that the computational efficiency of the SPRS is not only far superior to that of traditional Pawlak RS (TPRS), but also higher than that of most existing RSs.