Rough set attributes reduction based on adaptive PBIL algorithm

This paper presents a PBIL algorithm based on adaptive theory-giving that the traditional reduction of rough set is not unique and the process lasts for a long time. The learn probability and mutation rate of traditional PBIL algorithm can change adaptively by introducing the Systemic Entropy, then a self-learning and adaptive variability PBIL algorithm (APBIL) is formed. When it is applied to attributes reduction of rough set, it not only maintains the characteristics of global optimization but also reduces the correlation among attributes. Finally, the simplicity and effectiveness of the algorithm are demonstrated by an example.