Access the full text.
Sign up today, get DeepDyve free for 14 days.
Stephen Smith, Matthew Streeter (2007)
Using online algorithms to solve np-hard problems more efficiently in practice
Tzur Sayag, Shai Fine, Y. Mansour (2006)
Combining Multiple Heuristics
M. Gagliolo, C. Legrand, M. Birattari (2009)
Mixed-Effects Modeling of Optimisation Algorithm Performance
Lin Xu, F. Hutter, H. Hoos, Kevin Leyton-Brown (2007)
: The Design and Analysis of an Algorithm Portfolio for SAT
M. Luby, A. Sinclair, David Zuckerman (1993)
Optimal speedup of Las Vegas algorithms[1993] The 2nd Israel Symposium on Theory and Computing Systems
M. Gagliolo, J. Schmidhuber (2006)
Learning dynamic algorithm portfoliosAnnals of Mathematics and Artificial Intelligence, 47
N. Cesa-Bianchi, Y. Mansour, Gilles Stoltz (2005)
Improved second-order bounds for prediction with expert adviceMachine Learning, 66
N. Cesa-Bianchi, G. Lugosi, Gilles Stoltz (2004)
Minimizing regret with label efficient predictionIEEE Transactions on Information Theory, 51
L. Finkelstein, Shaul Markovitch, E. Rivlin (2003)
Optimal Schedules for Parallelizing Anytime Algorithms: The Case of Shared ResourcesJ. Artif. Intell. Res., 19
C. Gomes, B. Selman (2001)
Algorithm portfoliosArtif. Intell., 126
T. Schaul, Jürgen Schmidhuber (2010)
MetalearningScholarpedia, 5
M. Gagliolo, J. Schmidhuber (2008)
Towards Distributed Algorithm Portfolios
N. Littlestone, Manfred Warmuth (1989)
The weighted majority algorithm30th Annual Symposium on Foundations of Computer Science
Matthew Streeter, D. Golovin, Stephen Smith (2007)
Combining Multiple Heuristics Online
M. Gagliolo, Viktor Zhumatiy, J. Schmidhuber (2004)
Adaptive Online Time Allocation to Search Algorithms
E. Horvitz, S. Zilberstein (2001)
Computational tradeoffs under bounded resourcesArtif. Intell., 126
J. Kolen (1988)
Faster learning through a probabilistic approximation algorithmIEEE 1988 International Conference on Neural Networks
H. Robbins (1952)
Some aspects of the sequential design of experimentsBulletin of the American Mathematical Society, 58
M. Muselli, M. Rabbia (1991)
Parallel trials versus single search in supervised learning
Chamy Allenberg, P. Auer, L. Györfi, György Ottucsák (2006)
Hannan Consistency in On-Line Learning in Case of Unbounded Losses Under Partial Monitoring
Daniel Commenges (1998)
Book Review: Survival Analysis: Techniques for Censored and Truncated Data. John P. Klein and Melvin L. Moeschberger, Springer Verlag, 1997. No. of pages: 502. Price: DM 94. ISBN 0-387-94829-5Statistics in Medicine, 17
R. Battiti, M. Brunato, Franco Mascia (2008)
Reactive Search and Intelligent Optimization
J. Corchado, S. Rodríguez, J. Llinas, J. Molina (2009)
International Symposium on Distributed Computing and Artificial Intelligence, DCAI 2008, University of Salamanca, Spain, 22th-24th October 2008, 50
Marek Petrik (2004)
Statistically Optimal Combination of Algorithms
JP Klein, ML Moeschberger (2003)
Survival Analysis: Techniques for Censored and Truncated Data
Matthew Streeter, Stephen Smith (2008)
New Techniques for Algorithm Portfolio Design
M. Gagliolo, C. Legrand (2010)
Algorithm Survival Analysis
F. Hutter, Y. Hamadi (2005)
Parameter Adjustment Based on Performance Prediction: Towards an Instance-Aware Problem Solver
H. Hoos, T. Stützle (2004)
Stochastic Local Search: Foundations & Applications
C. Gomes, B. Selman, N. Crato, Henry Kautz (2000)
Heavy-Tailed Phenomena in Satisfiability and Constraint Satisfaction ProblemsJournal of Automated Reasoning, 24
M. Gagliolo, J. Schmidhuber (2008)
Algorithm Selection as a Bandit Problem with Unbounded LossesArXiv, abs/0807.1494
Lin Xu, F. Hutter, H. Hoos, Kevin Leyton-Brown (2008)
SATzilla: Portfolio-based Algorithm Selection for SATJ. Artif. Intell. Res., 32
L. Finkelstein, Shaul Markovitch, E. Rivlin (2002)
Optimal schedules for parallelizing anytime algorithms: the case of independent processes
(1997)
Machine Learning. McGraw-Hill Science/Engineering/Math
R. Dromey (1986)
An algorithm for the selection problemSoftware: Practice and Experience, 16
M Gagliolo, J Schmidhuber (2010)
Learning and Intelligent Optimization. 4th International Conference, LION 4, Venice, Italy, January 18–22 (2010) Selected Papers
G. Molenberghs, G. Verbeke (2008)
Longitudinal data analysis
EJ Horvitz, S Zilberstein (2001)
Computational tradeoffs under bounded resources (editorial)Artif. Intell., 126
EL Kaplan, P Meyer (1958)
Nonparametric estimation from incomplete samplesJ. Am. Stat. Assoc., 73
(2004)
Machine Learning : ECML 2004 . Proceedings of the 15 th European Conference on Machine Learning
M. Gagliolo, J. Schmidhuber (2007)
Learning Restart Strategies
E. Kaplan, P. Meier (1958)
Nonparametric Estimation from Incomplete ObservationsJournal of the American Statistical Association, 53
R Battiti, M Brunato, F Mascia (2008)
Operations Research/Computer Science Interfaces, vol. 45
Jens Bleiholder, Felix Naumann (2009)
Data fusionACM Comput. Surv., 41
Eugene Nudelman, Kevin Leyton-Brown, H. Hoos, Alex Devkar, Y. Shoham (2004)
Understanding Random SAT: Beyond the Clauses-to-Variables Ratio
M. Puterman (1994)
Markov Decision Processes: Discrete Stochastic Dynamic Programming
T. Bartz-Beielstein, Marco Chiarandini, Lus Paquete, M. Preuss (2010)
Experimental Methods for the Analysis of Optimization Algorithms
L. Babai (2006)
Monte-Carlo algorithms in graph isomorphism testing
K. Smith‐Miles (2009)
Cross-disciplinary perspectives on meta-learning for algorithm selectionACM Comput. Surv., 41
M. Gagliolo (2007)
Universal searchScholarpedia, 2
B. Huberman, R. Lukose, T. Hogg (1997)
An Economics Approach to Hard Computational ProblemsScience, 275
M Gagliolo, C Legrand (2010)
Empirical Methods for the Analysis of Optimization Algorithms, Natural Computing
P. Auer, N. Cesa-Bianchi, Y. Freund, R. Schapire (2002)
The Nonstochastic Multiarmed Bandit ProblemSIAM J. Comput., 32
N Littlestone, MK Warmuth (1994)
The weighted majority algorithmInf. Comput., 108
M Gagliolo, J Schmidhuber (2005)
Artificial Neural Networks: Formal Models and Their Applications—ICANN, Proceedings, Part 2
M. Gagliolo (2010)
Online Dynamic Algorithm Portfolios
M. Gagliolo, J. Schmidhuber (2005)
A Neural Network Model for Inter-problem Adaptive Online Time Allocation
Marek Petrik, S. Zilberstein (2006)
Learning parallel portfolios of algorithmsAnnals of Mathematics and Artificial Intelligence, 48
Kevin Leyton-Brown, Eugene Nudelman, Y. Shoham (2002)
Learning the Empirical Hardness of Optimization Problems: The Case of Combinatorial Auctions
We propose a method that learns to allocate computation time to a given set of algorithms, of unknown performance, with the aim of solving a given sequence of problem instances in a minimum time. Analogous meta-learning techniques are typically based on models of algorithm performance, learned during a separate offline training sequence, which can be prohibitively expensive. We adopt instead an online approach, named GAMBLETA, in which algorithm performance models are iteratively updated, and used to guide allocation on a sequence of problem instances. GAMBLETA is a general method for selecting among two or more alternative algorithm portfolios. Each portfolio has its own way of allocating computation time to the available algorithms, possibly based on performance models, in which case its performance is expected to improve over time, as more runtime data becomes available. The resulting exploration-exploitation trade-off is represented as a bandit problem. In our previous work, the algorithms corresponded to the arms of the bandit, and allocations evaluated by the different portfolios were mixed, using a solver for the bandit problem with expert advice, but this required the setting of an arbitrary bound on algorithm runtimes, invalidating the optimal regret of the solver. In this paper, we propose a simpler version of GAMBLETA, in which the allocators correspond to the arms, such that a single portfolio is selected for each instance. The selection is represented as a bandit problem with partial information, and an unknown bound on losses. We devise a solver for this game, proving a bound on its expected regret. We present experiments based on results from several solver competitions, in various domains, comparing GAMBLETA with another online method.
Annals of Mathematics and Artificial Intelligence – Springer Journals
Published: Apr 1, 2011
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.