Abstract
A popular way of dealing with the complexity of learning from examples is to proceed in an example-driven fashion. In the past, several researchers have shown that using an example-driven approach, it is possible to learn even structurally complex generalizations which would have been difficult to find using other multirelational learning (ILP) algorithms. On the other hand, it is also well known that the quality of the learning results in example-driven learning may depend on the ordering of the examples; however, such stability issues have received almost no attention. In this paper, we present empirical results in several multirelational application domains to show that instability actually affects the performance of a well-known example-driven ILP system. At the same time, we examine one possible solution to the instability problem, presenting an algorithm which relies on stochastically selected examples and parallel search. We show that our algorithm almost eliminates the instability of example-driven search with limited additional effort.
This is a preview of subscription content, log in via an institution to check access.
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
Peter Clark and Tim Niblett. The CN2 induction algorithm. Machine Learning, 3(4):261–283, 1989.
Hiroshi Fujita, Naoki Yagi, Tomonobu Ozaki, and Koichi Furukawa. A new design and implementation of Progol by bottom-up computation. In S. Muggleton, editor, Inductive Logic Programming, volume 1314 of Lecture Notes in Artificial Intelligence, pages 163–174. Springer-Verlag, 1996.
Richard Kaye. Minesweeper is NP-complete. The Mathematical Intelligencer, 22(2):9–15, Spring 2000.
Tohgoroh Matsui, Nobuhiro Inuzuka, Hirohisa Seki, and Hidenori Itoh. Comparison of three parallel implementations of an induction algorithm. In Proceedings of 8th International Parallel Computing Workshop, pages 181–188, Singapore, 1998.
Ryszard S. Michalski. On the quasi-minimal solution of the general covering problem. In Proceedings of the V International Symposium on Information Processing (FCIP 69), volume A3 (Switching Circuits), pages 125–128, Yugoslavia, Bled, 1969.
Tom M. Mitchell. Machine Learning. McGraw Hill, first edition, 1997.
Stephen Muggleton and Wray L. Buntine. Machine invention of first-order predicates by inverting resolution. In J. Laird, editor, Proceedings of the 5th Conference on Machine Learning, pages 339–352, San Mateo, CA, 1988. Morgan Kaufmann.
Stephen Muggleton and Cao Feng. Efficient induction of logic programs. In Stephen Muggleton, editor, Inductive Logic Programming, volume 38 of APIC Series, pages 281–298. Academic Press, London, 1992.
Stephen Muggleton and John Firth. Relational rule induction with CProgol4.4: a tutorial introduction. In Sašo Džeroski and Nada Lavrac, editors, Relational Data Mining, pages 160–187. Springer-Verlag, 2001.
Stephen Muggleton. Inverse entailment and Progol. New Generation Computing Journal, 13:245–286, 1995.
Shan-Hwei Nienhuys-Cheng and Ronald de Wolf. Foundations of Inductive Logic Programming, volume 1228 of Lecture Notes in Artificial Intelligence. Springer-Verlag, Berlin, first edition, February 1997.
J.Ross Quinlan and R. Michael Cameron-Jones. Introduction of logic programs: FOIL and related systems. New Generation Computing, Special issue on Inductive Logic Programming, 13(3–4):287–312, 1995.
Ashwin Srinivasan, Ross D. King, and Stephen Muggleton. The role of background knowledge: using a problem from chemistry to examine the performance of an ILP program. Transactions on Knowledge and Data Engineering, 1999. under review.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Peña Castillo, L., Wrobel, S. (2002). On the Stability of Example-Driven Learning Systems: A Case Study in Multirelational Learning. In: Coello Coello, C.A., de Albornoz, A., Sucar, L.E., Battistutti, O.C. (eds) MICAI 2002: Advances in Artificial Intelligence. MICAI 2002. Lecture Notes in Computer Science(), vol 2313. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-46016-0_34
Download citation
DOI: https://doi.org/10.1007/3-540-46016-0_34
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-43475-7
Online ISBN: 978-3-540-46016-9
eBook Packages: Springer Book Archive