Probability density function approximation using fuzzy rules

In this work, we present how to encode a probability density function (pdf) by using fuzzy logic. Our approach saves storage space when compared to approaches encoding the pdf in a lookup table. In addition, we do not need any interpolation scheme because the universal approximator capabilities of fuzzy systems. We present a procedure to optimally encode pdf values for unidimensional and bidimensional pdfs. We also evaluate the performance of our system with respect to the number of fuzzy sets used to describe pdfs in a symbolic way.