A multi-objective approach to robust optimization over time considering switching cost

Huang, Yuanjun; Ding, Yongsheng; Hao, Kuangrong; Jin, Yaochu

doi:10.1016/j.ins.2017.02.029

Cited by 40 publications

(34 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GDBG provided six properties of the environmental dynamics including small step change, large step change, random change, recurrent change, recurrent change with noise, and chaotic change. These environmental dynamics were used in some other studies such as [72], [73], in which the width and height of each peak changed using them.…”

Section: Dop Benchmarksmentioning

confidence: 99%

Scaling Up Dynamic Optimization Problems: A Divide-and-Conquer Approach

Yazdani

Omidvar

Branke

et al. 2020

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

Scalability is a crucial aspect of designing efficient algorithms. Despite their prevalence, large-scale dynamic optimization problems are not well-studied in the literature. This paper is concerned with designing benchmarks and frameworks for the study of large-scale dynamic optimization problems. We start by a formal analysis of the moving peaks benchmark and show its nonseparable nature irrespective of its number of peaks. We then propose a composite moving peaks benchmark suite with exploitable modularity covering a wide range of scalable partially separable functions suitable for the study of largescale dynamic optimization problems. The benchmark exhibits modularity, heterogeneity, and imbalance features to resemble real-world problems. To deal with the intricacies of large-scale dynamic optimization problems, we propose a decompositionbased coevolutionary framework which breaks a large-scale dynamic optimization problem into a set of lower dimensional components. A novel aspect of the framework is its efficient bilevel resource allocation mechanism which controls the budget assignment to components and the populations responsible for tracking multiple moving optima. Based on a comprehensive empirical study on a wide range of large-scale dynamic optimization problems with up to 200 dimensions, we show the crucial role of problem decomposition and resource allocation in dealing with these problems. The experimental results clearly show the superiority of the proposed framework over three other approaches in solving large-scale dynamic optimization problems.

show abstract

Section: Dop Benchmarksmentioning

confidence: 99%

Scaling Up Dynamic Optimization Problems: A Divide-and-Conquer Approach

Yazdani

Omidvar

Branke

et al. 2020

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

show abstract

“…In [10], a new two-layer multi-objective method was proposed to find robust solutions that can maximize both survival time and average fitness. In [11], another multiobjective method was proposed to minimize switching cost and maximize survival time. A PSO algorithm was used as the optimizer.…”

Section: A Robust Optimization Over Timementioning

confidence: 99%

“…Switching cost is Euclidean distance between robust solutions in successive environments. All of the proposed methods in [6], [8], [9] and [11] used predicted future fitness values of solutions for selecting robust solutions. In [6], an RBFN was used for approximating previous fitness values of solutions and an AR was used for predicting future values.…”

Section: A Robust Optimization Over Timementioning

confidence: 99%

Robust Optimization Over Time by Learning Problem Space Characteristics

Yazdani

Nguyễn

Branke

2019

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

show abstract

“…It is a hot topic in evolutionary computation. Some previously published methods on robust evolutionary optimization have been presented in [18][19][20][21][22][23][24][25]. In [18], the method of adopting the average fitness value of an individual's neighbors instead of the fitness value of that individual.…”

Section: Introductionmentioning

confidence: 99%

“…However, reduced computational cost is a less important factor to consider. A generic multi-objective optimization framework for robust optimization over time that simultaneously maximizes the robustness and minimizes the switching cost was proposed in [25]. The predicted fitness of the current solutions in a future environment according to their fitness values in the current and previous, based on which the predictor can be constructed, however, predictor construction also need additional time in the paper.…”

Section: Introductionmentioning

confidence: 99%