Pages: 255 - 264
Abstract:
Mathematical foundations are steadily extended and pushing rough set theory into incorporating new data analysis methods and data models. Generalized approximation spaces present abstract model useful in understanding unknown and undefined data structure leading into creation many new robust and intelligent approaches. Covering approximation spaces present data by means of coverings of the universe. In the paper, these two approaches have been put together introducing the concept of generalized covering approximation space. Further rough coverings model for generalized covering approximation spaces has been presented. Proposed rough covering models are based upon clustering and thresholding of feature space, are embedded in generalized approximation spaces, simultaneously spanning standard, fuzzy and probabilistic data models.
Keywords
generalized approximation spaces, covering approximation spaces, rough sets, fuzzy sets, probabilistic sets
Full article:
str_255-264_kwartalnik_nr24.pdf