Computational Scientific Discovery

Computational scientific discovery is becoming increasingly important in many areas of science. This chapter reviews the application of computational methods in the formulation of scientific ideas, that is, in the characterization of phenomena and the generation of scientific explanations, in the form of hypotheses, theories, and models. After a discussion of the evolutionary and anthropological roots of scientific discovery, the nature of scientific discovery is considered, and an outline is given of the forms that scientific discovery can take: direct observational discovery, finding empirical rules, and discovery of theories. A discussion of the psychology of scientific discovery includes an assessment of the role of induction. Computational discovery methods in mathematics are then described. This is followed by a survey of methods and associated applications in computational scientific discovery, covering massive systematic search within a defined space; rule-based reasoning systems; classification, machine vision, and related techniques; data mining; finding networks; evolutionary computation; and automation of scientific experiments. We conclude with a discussion of the future of computational scientific discovery, with consideration of the extent to which scientific discovery will continue to require human input.