Abstract
Exact learning of half-spaces over finite subsets of ℝn from membership queries is considered. We describe the minimum set of labelled examples separating the target concept from all the other ones of the concept class under consideration. For a domain consisting of all integer points of some polytope we give non-trivial lower bounds on the complexity of exact identification of half-spaces. These bounds are near to known upper bounds.
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
Angluin, D.: Queries and concept learning. Machine Learning (2) (1988) 319–342
Anthony, M., Brightwell, G., Shawe-Taylor, J.: On specifying Boolean functions by labelled examples. Discrete Applied Mathematics 61(1) (1995) 1–25
Bárány, I., Howe, R., Lovász, L.: On integer points in polyhedra: a lower bound. Combinatorica (12) (1992) 135–142
Bultman, W. J., Maass, W.: Fast identification of geometric objects with membership queries. Information and Computation 118(1) (1995) 48–64
Chernikov, S. N.: Linear Inequalities. “Nauka” Moscow (1968). German transl.: VEB Deutscher Verlag Wiss. Berlin (1971)
Chirkov, A. Yu.: On lower bound of the number of vertices of a convex hull of integer and partially integer points of a polyhedron. Proceedings of the First International Conference “Mathematical Algorithms”. NNSU Publishers Nizhny Novgorod (1995) 128–134 (Russian)
Hegedüs, T.: Geometrical concept learning and convex polytopes. Proceedings of the 7th Annual ACM Conference on Computational Learning Theory (COLT’94). ACM Press New York (1994) 228–236
Hegedüs, T.: Generalized teaching dimensions and the query complexity of learning. Proceedings of the 8th Annual ACM Conference on Computational Learning Theory (COLT’95). AM Press New York (1995) 108–117
Korobkov, V. K.: On monotone functions of logic algebra. Cybernetics Problems. ℋauka” Moscow 13 (1965) 5–28 (Russian)
Maass, W, Turán, Gy.: Lower bound methods and separation results for on-line learning models. Machine Learning (9) (1992) 107–145
Moshkov, M. Yu.: Conditional tests. Cybernetics Problems. ℋauka” Moscow 40 (1983) 131–170 (Russian)
Schrijver, A.: Theory of Linear and Integer Programming. Wley-Interscience New York (1986)
Shevchenko, V. N.: On some functions of many-valued logic connected with integer programming. Methods of Discrete Analysis in the Theory of Graphs and Circuits. Novosibirsk 42 (1985) 99–102 (Russian)
Shevchenko, V. N.: Deciphering of a threshold function of many-valued logic. Combinatorial-Algebraic Methods in Applied Mathematics. Gorky (1987) 155–163 (Russian)
Shevchenko, V. N.: Qualitative Topics in Integer Linear Programming. ▾izmatlit” Moscow (1995). English transl.: AMS Providence Rhode Island (1997)
Shevchenko, V. N., Zolotykh, N. Yu.: Decoding of threshold functions defined on the integer points of a polytope. Pattern Recognition and Image Analysis. MAIK/Interperiodica Publishing Moscow 7(2) (1997) 235–240
Shevchenko, V. N., Zolotykh, N. Yu.: On complexity of deciphering threshold functions of k-valued logic. Russian Math. Dokl. (Doklady Rossiiskoi Akademii Nauk) (to appear)
Veselov, S. I.: A lower bound for the mean number of irreducible and extreme points in two discrete programming problems. Manuscript No. 619-84, deposited at VINITI Moscow (1984) (Russian).
Zolotykh, N. Yu.: An algorithm of deciphering a threshold function of k-valued logic in the plane with the number of calls to the oracle O(log k). Proceedings of the First International Conference “Mathematical Algorithms”. NNSU Publishers Nizhny Novgorod (1995) 21–26 (Russian)
Zolotykh, N. Yu., Shevchenko, V. N.: On complexity of deciphering threshold functions. Discrete Analysis and Operations Research. Novosibirsk 2(1) (1995) 72–73 (Russian)
Zolotykh, N. Yu., Shevchenko, V. N.: Deciphering threshold functions of k-valued logic. Discrete Analysis and Operations Research. Novosibirsk 2(3) (1995) 18–23. English transl.: Korshunov, A. D. (ed.): Operations Research and Discrete Analysis. Kluwer Ac. Publ. Netherlands (1997) 321-326
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
Shevchenko, V.N., Zolotykh, N.Y. (1998). Lower Bounds for the Complexity of Learning Half-Spaces with Membership Queries. 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_5
Download citation
DOI: https://doi.org/10.1007/3-540-49730-7_5
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