Abstract
Theories are collections of large bodies of data in the real world. We describe autonomous systems, which observe the outside world and try to generate programs which reproduce the observed data. Methods for generation of new programs are enumeration as well as mutation and combination of old programs. We describe two criteria for judging the quality of a program. We can judge a program to be good if it is short and describes a large body of input data. With this criterion we show that a system can learn to evaluate arithmetic expressions in polish notation. But we can also judge a program to be good if it allows to compress the total length of descriptions ofall observations so far. By the latter criterion a system can createtests which can be used e.g. to partition the programs found so far into directories.
Similar content being viewed by others
References
A. Birk, Beschreibung und Implementierung eines lernenden Systems. Master's thesis, Universität Saarbrücken (1993).
P. Bergmann, J. Keller, T. Malter, S.M. Müller, W.J. Paul, T. Pöschel, O. Schlüter and L. Thiele, Implementierung eines informationstheoretischen Ansatzes zur Bilderkennung, in:Innovative Informations-Infrastrukturen, III-Forum, eds. B. Gollan, W.J. Paul, and A. Schmitt (Springer, 1988) pp 187–197.
P. Bergmann, W.J. Paul and L. Thiele, An information theoretic approach to computer vision, in:Dynamical Networks, eds. W. Ebeling and M. Peschel (Akademie-Verlag Berlin, 1989) pp. 52–58.
J.G. Carbonell, Paradigms for machine learning, in:Machine Learning Paradigms and Methods, ed. J. Carbonell (MIT Elsevier, 1989) pp. 1–9.
E.M. Gold, Language identification in the limit. Inf. Contr. 10 (1967) 447–474.
D. Haussler, Quantifying inductive bias: AI learning algorithms and Valiant's learning framework, Art. Int. 36 (1988) 177–221.
J.R. Koza,Genetic Programming (The MIT Press, 1992).
L.A. Levin, Universal search problems, Probl. Inf. Transmission 9 (1973) 265–266.
S. Minton, J.G. Carbonell, C.A. Knoblock, D.R. Kuokka, O.Etzioni and Y.Gil, Explanation-based learning: A problem solving perspective, in:Machine Learning Paradigms and Methods, ed. J. Carbonell (MIT Elsevier, 1989) pp. 63–118.
B.K. Natarajan,Machine Learning (Morgan Kaufmann, 1991).
J. Keller P. Bergmann and W.J. Paul, A selforganizing system for image recognition, in:Machine Learning and Neural Networks, ed. M.H. Hamza (IASTED, Acta Press, Anaheim, Calgary, Zürich, 1990) pp. 33–36.
H. Ritter, T. Martinez and K. Schulten,Neuronale Netze (Addison Wesley, 1991).
R.J. Solomonoff, Complexity-based induction systems: Comparisons and convergence theorems, IEEE Trans. Inf. Transmission IT-24 (1978).
R.J. Solomonoff, Optimum sequential search, Technical report, Oxbridge Research, Box 559, Cambridge, MA 02238 (1985).
R.J. Solomonoff, A system for incremental learning based on algorithmic probability, in:Proc. 6th Israeli Conf. on Artificial Intelligence, Computer Vision and Pattern Recognition (1989) pp. 515–527.
L.G. Valiant, A theory of the learnable, Commun. ACM 27 (1984) 1134–1142.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Paul, W.J., Solomonoff, R. Autonomous theory building systems. Ann Oper Res 55, 179–193 (1995). https://doi.org/10.1007/BF02031720
Issue Date:
DOI: https://doi.org/10.1007/BF02031720