Boolean Function Complementation Based Algorithm for Data Discretization

This paper presents a fast algorithm for discretization of decision tables. An important novelty of the proposed solution is the application of the original algorithm of Boolean function complementation, which is a basic procedure of the field of logic synthesis, in the process of discretizing the data. This procedure has already been used by the author to calculate reducts of decision tables, where the time of calculation has been significantly reduced. It yields the idea of using the algorithm of complementation in the process of discretization. The algorithm has been generalized for the discretization of inconsistent decision tables and is used in the processing of numerical data from various fields of technology, especially for multimedia data.