The role of computational intelligence in data mining

In this paper, we explore the interdependent roles of computational intelligence, data mining, and machine learning. We explain how these three seemingly independent AI specialties are inextricably linked through the science of randomization and reuse. Computational intelligence is shown to imply soft computing, whose bounds are soft or fuzzy only because they need be adoptively formed (e.g., type II fuzzy logic). Data mining implies the compression of data (including anomalous data) to its simplest form. Such randomization operations are inherently incomplete and thus are necessarily domain-specific or adaptive as a consequence. Finally, machine learning is the study of computational adaptation. For example, a machine that plays master-level chess is not necessarily computationally intelligent - unless it is capable of adaptive improvement in its play. This paper will attempt to make it clear that intelligence and the capability for learning are in fact synonyms for an AI.