Tuning Metaheuristics - A Machine Learning Perspective

By Mauro Birattari. Studies in Computational Intelligence, 2009

348 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/e63f7fc2e8fb6a04a39ccc5e4ebe5decd62263b9

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Automated Configuration of Mixed Integer Programming Solvers (2010, methodology): “Some relatively recent model-free approaches permit both larger numbers of parameters and categorical domains, in particular Composer [12], F-Race [9, 8], GGA [3], and our own ParamILS [20, 19].”
Automatically improving the anytime behaviour of optimisation algorithms (2014, methodology): “…the performance of optimisation algorithms, when designing and evaluating new algorithms and when tuning existing algorithms to specific problems (Birattari, 2009; Bartz-Beielstein, 2006; Hutter, Hoos, Leyton-Brown, & Stützle, 2009; Eiben & Smit, 2011; Hutter, Hoos, & Leyton-Brown, 2011; Hoos,…”
A Survey of Automatic Parameter Tuning Methods for Metaheuristics (2020, methodology): “F-Race, which is inspired from the algorithm Hoeffding race [49], [50] in machine learning for model selection, was proposed in [45] and comprehensively studied in [31].”
A survey on optimization metaheuristics (2013, background): “Recent research reveals that the tuning problem has much in common with the problems that are typically faced in machine learning [27].”
Fast cross-validation via sequential testing (2012, methodology): “Birattari et al. (2002) apply racing in the domain of evolutionary algorithms and extend the framework by using the Friedman test to filter out non-promising configurations. While Bradley and Schapire (2008) use similar concepts in the context of boosting (FilterBoost), Mnih et al. (2008) introduce ”
The irace package: Iterated racing for automatic algorithm configuration (2016, methodology): “It also implements several extensions already described by Birattari [19] , such as the use of the paired t -test instead of Friedman’s test.”
F-Race and Iterated F-Race: An Overview (2010, methodology): “Assuming that the performance of a stochastic algorithm is evaluated by a total of N runs, it has been proved by Birattari (2004a, 2009) that sampling N instances with one run on each instance results in the lowest variance of the estimator.”
Hybrid metaheuristics in combinatorial optimization: A survey (2011, background): “Furthermore, discussions on the overall experimental methodology or just one of its issues, such as parameter tuning or the statistical assessment of results, can be found in [167, 164, 168, 169].”