Active Learning of Group-Structured Environments

  • Gábor Bartók
  • Csaba Szepesvári
  • Sandra Zilles
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5254)


The question investigated in this paper is to what extent an input representation influences the success of learning, in particular from the point of view of analyzing agents that can interact with their environment. We investigate learning environments that have a group structure. We introduce a learning model in different variants and study under which circumstances group structures can be learned efficiently from experimenting with group generators (actions). Negative results are presented, even without efficiency constraints, for rather general classes of groups showing that even with group structure, learning an environment from partial information is far from trivial. However, positive results for special subclasses of Abelian groups turn out to be a good starting point for the design of efficient learning algorithms based on structured representations.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Babai, L., Fried, K.: Approximate representation theory of finite groups. In: Proc. 32nd Annual Symposium on Foundations of Computer Science, pp. 733–742 (1991)Google Scholar
  2. 2.
    Babai, L., Szemerédi, E.: On the complexity of matrix group problems I. In: IEEE Symposium on Foundations of Computer Science (1984)Google Scholar
  3. 3.
    Boone, W.W.: The Word Problem. The Annals of Mathematics 70, 207–265 (1959)CrossRefMathSciNetGoogle Scholar
  4. 4.
    Culberson, J.C., Schaeffer, J.: Pattern databases. Computational Intelligence 14, 318–334 (1998)CrossRefMathSciNetGoogle Scholar
  5. 5.
    Friedl, K., Ivanyos, G., Santha, M.: Efficient testing of groups. In: Proc. 37th Annual ACM Symposium on Theory of Computing, pp. 157–166 (2005)Google Scholar
  6. 6.
    Gold, E.M.: Language identification in the limit. Inform. Control 10, 447–474 (1967)MATHCrossRefGoogle Scholar
  7. 7.
    Holte, R., Grajkowski, J., Tanner, B.: Hierarchical heuristic search revisited. In: Symposium on Abstraction, Reformulation and Approximation (2005)Google Scholar
  8. 8.
    Jaeger, H.: Observable operator models for discrete stochastic time series. Neural Computation 12, 1371–1398 (2000)CrossRefGoogle Scholar
  9. 9.
    Korf, R.E.: Finding optimal solutions to Rubik’s cube using pattern databases. In: AAAI/IAAI, pp. 700–705 (1997)Google Scholar
  10. 10.
    Korf, R.E.: Analyzing the performance of pattern database heuristics. In: Proc. 22nd AAAI Conference on Artificial Intelligence, pp. 1164–1170 (2007)Google Scholar
  11. 11.
    Littman, M.L., Sutton, R., Singh, S.: Predictive representations of state. In: Advances in Neural Information Processing Systems 14, pp. 1555–1561 (2002)Google Scholar
  12. 12.
    Novikov, P.S.: On the algorithmic undecidability of the word problem in group theory. In: Proc. Steklov Institute of Mathematics, vol. 44, pp. 1–143 (1955) (in Russian)Google Scholar
  13. 13.
    Rivest, R.L., Schapire, R.E.: Diversity-based inference of finite automata. J. ACM, 555–589 (1994)Google Scholar
  14. 14.
    Rothman, J.J.: An Introduction to the Theory of Groups. Springer, Heidelberg (1995)Google Scholar
  15. 15.
    Stephan, F., Ventsov, Y.: Learning algebraic structures from text. Theoret. Comput. Sci. 268(2), 221–273 (2001)MATHCrossRefMathSciNetGoogle Scholar
  16. 16.
    Strehl, A.L., Diuk, C., Littman, M.L.: Efficient structure learning in factored-state MDPs. In: Proc. 22nd AAAI Conference on Artificial Intelligence, pp. 645–650 (2007)Google Scholar
  17. 17.
    Vinodchandran, N.V.: Counting Complexity and Computational Group Theory. PhD thesis, Institute of Mathematical Sciences, Chennai, India (1999)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Gábor Bartók
    • 1
  • Csaba Szepesvári
    • 1
  • Sandra Zilles
    • 1
  1. 1.Department of Computing ScienceUniversity of AlbertaEdmontonCanada

Personalised recommendations