The dispersion metric and the CMA evolution strategy

Lunacek, Monte; Whitley, Darrell

doi:10.1145/1143997.1144085

Cited by 130 publications

(76 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The well-known Rastrigin function and the Ackley function belong to this class of landscapes. This observation led Lunacek and Whitley to introduce the dispersion metric [5] as statistical measure that attempts detecting such landscape topologies and may thus serve as a predictor for success or failure of CMA-ES. Both convex and globally convex landscapes are also termed "single-funnel landscapes" [5,6].…”

Section: Landscape Topologies and Their Impact On Continuous Black-bomentioning

confidence: 99%

Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis

Müller

Sbalzarini

2011

Applications of Evolutionary Computation

View full text Add to dashboard Cite

Abstract. We interpret real-valued black-box optimization problems over continuous domains as black-box landscapes. The performance of a given optimization heuristic on a given problem largely depends on the characteristics of the corresponding landscape. Designing statistical measures that can be used to classify landscapes and quantify their topographical properties is hence of great importance. We transfer the concept of fitness-distance analysis from theoretical biology and discrete combinatorial optimization to continuous optimization and assess its potential to characterize black-box landscapes. Using the CEC 2005 benchmark functions, we empirically test the robustness and accuracy of the resulting landscape characterization and illustrate the limitations of fitness-distance analysis. This provides a first step toward a classification of real-valued black-box landscapes over continuous domains.

show abstract

Section: Landscape Topologies and Their Impact On Continuous Black-bomentioning

confidence: 99%

Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis

Müller

Sbalzarini

2011

Applications of Evolutionary Computation

View full text Add to dashboard Cite

show abstract

“…In this context, the dispersion metric (dm) (Lunacek and Whitley 2006) is a technique to obtain information about the global structure of the landscape, by measuring the dispersion of good solutions. Ideally, if good solutions are very close, we might have a single funnel structure.…”

Section: Dispersion Metricmentioning

confidence: 99%

Dynamic selection of evolutionary operators based on online learning and fitness landscape analysis

et al. 2016

View full text Add to dashboard Cite

Self-adaptive mechanisms for the identification of the most suitable variation operator in evolutionary algorithms rely almost exclusively on the measurement of the fitness of the offspring, which may not be sufficient to assess the optimality of an operator (e.g., in a landscape with an high degree of neutrality). This paper proposes a novel adaptive operator selection mechanism which uses a set of four fitness landscape analysis techniques and an online learning algorithm, dynamic weighted majority, to provide more detailed information about the search space to better determine the most suitable crossover operator. Experimental analysis on the capacitated arc routing problem has demonstrated that different crossover operators behave differently during the search process, and selecting the proper one adaptively can lead to more promising results.

show abstract

“…In this context, the Dispersion Metric (dm) [17] is a technique to obtain information about the global structure of the landscape, by measuring the dispersion of the best solutions. Ideally, if the best solutions are very close we might be in presence a single funnel structure.…”

Section: Online Fitness Landscape Analysismentioning

confidence: 99%

Dynamic Selection of Evolutionary Algorithm Operators Based on Online Learning and Fitness Landscape Metrics

Consoli

Minku

Yao

2014

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Self-adaptive mechanisms for the identification of the most suitable variation operator in Evolutionary meta-heuristics rely almost exclusively on the measurement of the fitness of the offspring, which may not be sufficient to assess the optimality of an operator (e.g., in a landscape with an high degree of neutrality). This paper proposes a novel Adaptive Operator Selection mechanism which uses a set of four Fitness Landscape Analysis techniques and an online learning algorithm, Dynamic Weighted Majority, to provide more detailed informations about the search space in order to better determine the most suitable crossover operator on a set of Capacitated Arc Routing Problem (CARP) instances. Extensive comparison with a state of the art approach has proved that this technique is able to produce comparable results on the set of benchmark problems.

show abstract

The dispersion metric and the CMA evolution strategy

Cited by 130 publications

References 20 publications

Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis

Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis

Dynamic selection of evolutionary operators based on online learning and fitness landscape analysis

Dynamic Selection of Evolutionary Algorithm Operators Based on Online Learning and Fitness Landscape Metrics

Contact Info

Product

Resources

About