Computational Intelligence Techniques for Data Analysis

Abstract

The paper is a survey of the computational intelligence methods and their application to the data analysis problems. Neural networks, fuzzy sets, neuro-fuzzy systems, and genetic algorithms are considered. The advantages and disadvantages of the soft computing tools as well as specific issues of their application to data processing are analyzed, and the directions for their further improvement are outlined. New clustering algorithms that can operate under substantial uncertainty and cluster overlap are proposed.