Abstract
Randomized restart is an effective technique for eliminating heavy-tails and improving the performance of backtrack algorithms [1]. Different restart strategies use different cutoff schedules and some of the better studied ones include a fixed-cutoff strategy and Luby et al.’s universal strategy [2]. However, these strategies are more of theoretical interest and in practice Walsh’s geometric strategy seems to offer more tangible benefits [3]. Our two focuses are to firstly provide some theoretical results on the geometric strategy and secondly to establish an empirical method for studying the different strategies in a more systematic and efficient manner.
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
Similar content being viewed by others
References
Gomes, C.P., Selman, B., Crato, N., Kautz, H.: Heavy-tailed phenomenon in satis- fiability and constraint satisfaction problems. J. of Auto. Reas. 24, 67–100 (2000)
Luby, M., Sinclair, A., Zuckerman, D.: Optimal speedup of Las Vegas algorithms. Information Processing Letters 47, 173–180 (1993)
Walsh, T.: Search in a small world. In: Proceedings of IJCAI 1999, pp. 1172–1177 (1999)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Wu, H., van Beek, P. (2003). Restart Strategies: Analysis and Simulation. In: Rossi, F. (eds) Principles and Practice of Constraint Programming – CP 2003. CP 2003. Lecture Notes in Computer Science, vol 2833. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-45193-8_125
Download citation
DOI: https://doi.org/10.1007/978-3-540-45193-8_125
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-20202-8
Online ISBN: 978-3-540-45193-8
eBook Packages: Springer Book Archive