Abstract
An inductive-learning criterion is proposed for detection of regularities that describe classes of objects in the Boolean feature space. The criterion decides between hypotheses of random and regular occurrence of combinations of feature values in descriptions of objects in the learning sample.
Literature Cited
M. M. Bongard, Classification Problem [in Russian], Nauka, Moscow (1967).
N. D. Vashchenko, "Concept formation in a semantic network," Kibernetika, No. 2, 101–107 (1983).
V. P. Gladun, Heuristic Search in Complex Environments [in Russian], Naukova Dumka, Kiev (1977).
Yu. I. Zhuravlev, M. M. Kamilov, and Sh. E. Tulyaganov, Estimation Algorithms and Their Application [in Russian], FAN, Tashkent (1974).
E. Hunt, J. Marin, and F. Stone, Computer Modeling of the Concept Formation Process [Russian translation], Mir, Moscow (1970).
G. S. Lbov, Methods of Processing of Heterogeneous Experimental Data [in Russian], Nauka, Novosibirsk (1981).
A. D. Zakrevskii, "Detection of implicative regularities in a Boolean feature space and pattern recognition," Kibernetika, No. 1, 1–6 (1982).
A. D. Zakrevskii, Logic of Recognition [in Russian], Nauka i Tekhnika, Minsk (1988).
W. Feller, Introduction to Probability Theory and Its Applications, Vol. 2, Wiley, New York (1971).
Additional information
Translated from Kibernetika i Sistemnyi Analiz, No. 4, pp. 166–169, July–August, 1991.
Rights and permissions
About this article
Cite this article
Zavadskaya, L.A., Khomenko, L.V. Criterion for detection of regularities in a Boolean feature space. Cybern Syst Anal 27, 624–626 (1991). https://doi.org/10.1007/BF01130375
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01130375