Capping methods for the automatic configuration of optimization algorithms

Souza, Marcelo de; Ritt, Marcus; López-Ibáñez, Manuel

doi:10.1016/j.cor.2021.105615

Cited by 8 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The capping methods are implemented by the capopt package. The source code, instructions of use, examples and additional details can be found in De Souza, Ritt, and López-Ibáñez (2020). We applied the capping methods in the configuration of the following algorithms: ACOTSP (Dorigo and Stützle 2004), an ant colony optimization framework for the traveling salesperson problem; HEACOL (Galinier and Hao 1999;Lewis 2016), a hybrid evolutionary algorithm for the graph coloring; TSBPP Vigo 1999, 2004), a tabu search for the bin packing problem; HHBQP (De Souza and Ritt 2018), a hybrid heuristic for the unconstrained binary quadratic programming; LKH (Lin and Kernighan 1973;Helsgaun 2000Helsgaun , 2009Helsgaun , 2018, the Lin-Kernighan-Helsgaun algorithm for the traveling salesperson problem; and SCIP (Achterberg 2009), an open-source exact solver for mixed integer programming applied for solving the combinatorial auction winner determination problem.…”

Section: Resultsmentioning

confidence: 99%

How to Speed-Up the Automated Configuration of Optimization Algorithms

Souza

2021

SOCS

View full text Add to dashboard Cite

This paper presents a set of capping methods to speed-up the automated configuration of optimization algorithms. These methods use known previous executions to compute a performance envelope, which is used to evaluate new executions and early stop those with unsatisfactory performance. Preliminary experiments on six scenarios show that the capping methods save up to 78% of the configuration effort, while finding configurations of the same quality.

show abstract

Section: Resultsmentioning

confidence: 99%

How to Speed-Up the Automated Configuration of Optimization Algorithms

Souza

2021

SOCS

View full text Add to dashboard Cite

show abstract

“…All code, models and data are available in the repository accompanying this paper. 4 For the surrogates that use an informed prior, we use leave-one-out cross validation to split the models into prior and test set. To compare results across different models, we introduce a metric that denotes the relative improvement of the timeout.…”

Section: Methodsmentioning

confidence: 99%

“…However, to obtain these runtimes during the search, one is required to allow the solver to execute until completion, regardless of the quality of the configuration considered. This stands in stark contrast with adaptive capping, a technique for optimising runtime in which solver runs using unfruitful configurations are stopped from the moment the runtime surpasses that of the best configuration found so far [3,4,7].…”

Section: Introductionmentioning

confidence: 99%

Model-Based Algorithm Configuration with Adaptive Capping and Prior Distributions

Bleukx

Senne

Coenen

et al. 2022

Integration of Constraint Programming, Artificial Intelligence, and Operations Research

View full text Add to dashboard Cite

Many advanced solving algorithms for constraint programming problems are highly configurable. The research area of algorithm configuration investigates ways of automatically configuring these solvers in the best manner possible. In this paper, we specifically focus on algorithm configuration in which the objective is to decrease the time it takes the solver to find an optimal solution. In this setting, adaptive capping is a popular technique which reduces the overall runtime of the search for good configurations by adaptively setting the solver's timeout to the best runtime found so far. Additionally, sequential model-based optimization (SMBO)-in which one iteratively learns a surrogate model that can predict the runtime of unseen configurations-has proven to be a successful paradigm. Unfortunately, adaptive capping and SMBO have thus far remained incompatible, as in adaptive capping, one cannot observe the true runtime of runs that time out, precluding the typical use of SMBO. To marry adaptive capping and SMBO, we instead use SMBO to model the probability that a configuration will improve on the best runtime achieved so far, for which we propose several decomposed models. These models also allow defining prior probabilities for each hyperparameter. The experimental results show that our DeCaprio method speeds up hyperparameter search compared to random search and the seminal adaptive capping approach of ParamILS.

show abstract

“…The configurator may cancel runs according to some fixed cut off value or adaptively in relation to the runtime of other configurations. Although AC traditionally focused only on runtime capinng, recent work by De Souza et al (2021) introduces capping mechanisms that can be used with quality metrics. Terminating target algorithm runs before they finish results in censored information for the configurator where only a lower bound for the costs c(i, θ) is observed (Hutter et al, 2013;Eggensperger et al, 2018Eggensperger et al, , 2020.…”

Section: Configurator Viewmentioning

confidence: 99%

A Survey of Methods for Automated Algorithm Configuration

Schede¹,

Brandt²,

Tornede³

et al. 2022

Preprint

View full text Add to dashboard Cite

Algorithm configuration (AC) is concerned with the automated search of the most suitable parameter configuration of a parametrized algorithm. There is currently a wide variety of AC problem variants and methods proposed in the literature. Existing reviews do not take into account all derivatives of the AC problem, nor do they offer a complete classification scheme. To this end, we introduce taxonomies to describe the AC problem and features of configuration methods, respectively. We review existing AC literature within the lens of our taxonomies, outline relevant design choices of configuration approaches, contrast methods and problem variants against each other, and describe the state of AC in industry. Finally, our review provides researchers and practitioners with a look at future research directions in the field of AC.

show abstract

Capping methods for the automatic configuration of optimization algorithms

Cited by 8 publications

References 26 publications

How to Speed-Up the Automated Configuration of Optimization Algorithms

How to Speed-Up the Automated Configuration of Optimization Algorithms

Model-Based Algorithm Configuration with Adaptive Capping and Prior Distributions

A Survey of Methods for Automated Algorithm Configuration

Contact Info

Product

Resources

About