Abstract
This paper aims to extend well known results about polynomial update-time bounded learning strategies in a recursion theoretic setting.
We generalize the update-time complexity using the approach of Blum’s computational complexity measures.
It will turn out, that consistency is the first natural condition having a narrowing effect for arbitrary update boundaries. We show the existence of arbitrary hard, as well as an infinite chain of harder and harder consistent learnable sets. The complexity gap between consistent und inconsistent strategies solving the same problem can be arbitary large. We prove an exact characterization for polynomial consistent learnability, giving a deeper insight in the problem of hard consistent learnability.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
J. Barzdin. Inductive inference of automata, functions and programs. In Proceedings International Congress of Math., pages 455–460, Vancouver, 1974.
Balcazar, Diaz, Gavalda, and Watanabe. The query complexity of learning DFA. NEWGEN: New Generation Computing, 12, 1994.
M. Blum. A machine-independent theory of the complexity of recursive functions. In Journal of Association for Computing Machinery, volume 11, pages 322–336, April 1967.
R. P. Daley and C. H. Smith. On the complexity of inductive inference. In Information and Control, volume 69, pages 12–30, March 1986.
R. Freivalds, E. Kinber, and C.H. Smith. On the impact of forgetting on learning machines. Computer Science Technical Report Series CS-TR-3072, University of Maryland, College Park, MD, 20742, May 1993.
RŪsiņš Freivalds, E.m B. Kinber, and Carl H. Smith. On the intrinsic complexity of learning. Information and Computation, 123(1):64–71, 15 November 1995.
Michele Flammini. On the learnability of monotone kΜDNF formulae. Information Processing Letters, 52(3):167–173, 11 November 1994.
M. E. Gold. Limiting recursion. In Journal of Symbolic Logic, volume 30, pages 28–48, March 1965.
H. Ishizaka. Polynomial time learnability of simple deterministic languages. Machine Learning, 5(2):151–164, 1990. Special Issue on Computational Learning Theory; first appeared in 2nd COLT conference (1989).
K.P. Jantke and H.-R. Beick. Combining postulates of naturalness in inductive inference. In Elektronische Informationsverarbeitung und Kybernetik, volume 17, pages 465–484, 1984.
S. Jain and A. Sharma. The structure of intrinsic complexity of learning. Lecture Notes in Computer Science, 904, 1995.
E. Kinber. Monotonicity versus efficiency for learning languages from texts. Lecture Notes in Computer Science, 872, 1994.
S. Lange and R. Wiehagen. Polynomial-time inference of arbitrary pattern languages. In New Generation Computing, volume 8, pages 361–370, 1991.
S. Lange and T. Zeugmann. Trading monotonicity demands versus efficiency. Bulletin of Informatics and Cybernetics, 27:53–83, 1995.
L. Pitt. Inductive inference, DFAs and computational complexity. In Proceedings of the Workshop Analogical and Inductive Inference, volume 397 of LNAI, pages 18–44, 1989.
L. Pitt and M. K. Warmuth. Prediction preserving reducibility. J. of Comput. Syst. Sci., 41(3):430–467, December 1990. Special issue of the for the Third Annual Conference of Structure in Complexity Theory (Washington, DC., June 88).
R. Solomono. A formal theory of inductive inference. In Information and Control, volume 7, pages 1–22
L.G. Valiant. A theory of the learnable. In Comm. Assoc. Comp. Math., volume 27(11), pages 1134–1142, 1984.
Osamu Watanabe. A framework for polynomial-time query learnability. Mathematical Systems Theory, 27(3):211–229, May/June 1994.
R. Wiehagen and W. Liepe. Charakteristische Eigenschaften von erkennbaren Klassen rekursiver Funktionen. In Elektronische Informationsverarbeitung und Kybernetik, volume 12, pages 421–436, 1976.
R. Wiehagen and T. Zeugmann. Too much information can be too much for learning e.ciently. In 3rd Int. Workshop on Analogical and Inductive Inference, volume 642 of Lecture Notes in Artificial Intelligence, pages 72–86, 1992.
R. Wiehagen and T. Zeugamnn. Ignoring data may be the only way to learn efficiently. In Journal of Experimental and Theoretic Artifical Intelligence, volume 6, pages 131–144, 1994.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Stein, W. (1998). Consistent Polynomial Identification in the Limit. In: Richter, M.M., Smith, C.H., Wiehagen, R., Zeugmann, T. (eds) Algorithmic Learning Theory. ALT 1998. Lecture Notes in Computer Science(), vol 1501. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-49730-7_30
Download citation
DOI: https://doi.org/10.1007/3-540-49730-7_30
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65013-3
Online ISBN: 978-3-540-49730-1
eBook Packages: Springer Book Archive