Abstract
Reviewing structured weight-based prediction algorithms (SWP for short) due to Takimoto, Maruoka and Vovk, we present underlying design methods for constructing a variety of on-line prediction algorithms based on the SWP. In particular, we shown how the typical expert model where the experts are considered to be arranged on one layer can be generalized to the case where they are laid on a tree structure so that the expert model can be applied to search for the best pruning in a straightforward fashion through dynamic programming scheme.
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
N. Cesa-Bianchi, Y. Freund, D. Helmbold, D. Haussler, R. Schapire and M. Warmuth, How to use expert advice, JACM 44(3) (1997) 427–485.
Y. Freund and R. E. Schapire, A decision-theoretic generalization of on-line learning and an application to boosting, in: P. Vitányi, ed., Computational Learning Theory, Lecture Notes in Computer Science, Vol. 904 (Springer, Berlin, 1995) 23–37. To appear in J. Computer System Sciences.
D. Helmbold and R. Schapire, Predicting nearly as well as the best pruning of a decision tree, Machine Learning, 27 (1997) 51–68.
M. Kearns and Y. Mansour, A fast, bottom-up decision tree pruning algorithm with near-optimal generalization, To appear in the Machine Learning Conference.
N. Littlestone and M. K. Warmuth, The weighted majority algorithm, Inform. Computation 108 (1994) 212–261.
E. Takimoto, K. Hirai and A. Maruoka, A simple algorithm for predicting nearly as well as the best pruning labeled with the best prediction values of a decision tree, in: M. Li and A. Maruoka, eds., Algorithmic Learning Theory, Lecture Notes in Artificial Intelligence, Vol. 1316 (1997) 385–400.
E. Takimoto, A. Maruoka and V. Vovk, Predicting nearly as well as the best pruning of a decision tree through dynamic programming scheme, submitted.
V. Vovk, Aggregating strategies, in: Proc. 3rd COLT (Morgan Kaufmann, San Mateo, CA, 1990) 371–383.
V. Vovk, A game of prediction with expert advice, accepted for publication in J. Comput. Inform. Syst. Short version in: Proc. 8th COLT (Assoc. Comput. Mach., New York, 1995) 51–60.
L. G. Valiant, A theory of the learnable, Comm. ACM27 (1084) 1134–1142.
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
Maruoka, A., Takimoto, E. (1998). Structured Weight-Based Prediction Algorithms. 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_10
Download citation
DOI: https://doi.org/10.1007/3-540-49730-7_10
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