An inductive inference approach to classification
In this paper, we introduce a formal framework for investigating the relationship of inductive inference and the task of classification. We give the first results on the relationship between functions that can be identified in the limit and functions that can be acquired from unclassified objects only. Moreover, we present results on the complexity of classification functions and the preconditions necessary in order to allow the computation of such functions.
Unable to display preview. Download preview PDF.
- R. Freivalds, J. Barzdin, and K. Podnieks. Inductive inference of recursive functuions: complexity bounds. In Lecture Notes in Computer Science, volume 502, pages 111–151. Springer-Verlag, 1991.Google Scholar
- A. G. Hoffmann. Connectionist functionality and the emergent network behavior. Neurocomputing — An International Journal, 2(2):161–172, 1991.Google Scholar
- T. Kohonen. Self-Organization and Associative Memory. Springer-Verlag, Heidelberg, West-Germany, 1984.Google Scholar
- A. N. Kolmogorov. Three approaches to the quantitative definition of information. International Journal for Computer Mathematics, 2:157–168, 1968. (originally published in Russian: Problemi peredachi informacii, vol. 1, No. 1, 1965, p. 3–11.).Google Scholar
- L. Valiant. A theory of the learnable. Communications of the ACM, 27:1134–1142, 1984.Google Scholar
- C. von der Malsburg. Network self-organization. In S. F. Zornetzer, J. L. Davis, and C. Lau, editors, An Introduction to Neural and Electronic Networks, pages 421–432. Academic Press, New York, 1990.Google Scholar