An Indicator-Based Multiobjective Evolutionary Algorithm With Reference Point Adaptation for Better Versatility

Tian, Ye; Cheng, Ran; Zhang, Xingyi; Cheng, Fan; Jin, Yaochu

doi:10.1109/tevc.2017.2749619

Cited by 524 publications

(184 citation statements)

References 62 publications

Supporting

Mentioning

157

Contrasting

Order By: Relevance

“…The existing adaptation methods change the local density of reference points on the unit simplex. Such changes may not ensure the versatility and may also disturb the uniformity [27,28]. Moreover, the frequent periodic adaptations seem not appropriate, for they cannot judge the need for the adaptations, especially for the problems with full PFs where adaptations may push the population toward a part of the true PF [47,48].…”

Section: Motivationmentioning

confidence: 99%

“…Quite different from the dominance-based ideas, the indicator-based algorithms map the proximity and diversity of the population into designed indicators, such as hypervolume (HV) [23,24], averaged Hausdorff distance [25], R2-indicator [26] and IGD-NS indicator [27]. On one hand, researches have shown that the extensive compu-arXiv:1803.01097v4 [cs.NE] 20 Aug 2019 tation for the indicators in high-dimensional objective spaces still remains as a bottleneck [28], though some contributions have been made to alleviate the computational burden [29,30].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Many-Objective Evolutionary Algorithm With Two Interacting Processes: Cascade Clustering and Reference Point Incremental Learning

Zhao

Sun

et al. 2019

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

Researches have shown difficulties in obtaining proximity while maintaining diversity for many-objective optimization problems. Complexities of the true Pareto front pose challenges for the reference vector-based algorithms for their insufficient adaptability to the diverse characteristics with no priors. This paper proposes a many-objective optimization algorithm with two interacting processes: cascade clustering and reference point incremental learning (CLIA). In the population selection process based on cascade clustering (CC), using the reference vectors provided by the process based on incremental learning, the nondominated and the dominated individuals are clustered and sorted with different manners in a cascade style and are selected by round-robin for better proximity and diversity. In the reference vector adaptation process based on reference point incremental learning, using the feedbacks from the process based on CC, proper distribution of reference points is gradually obtained by incremental learning. Experimental studies on several benchmark problems show that CLIA is competitive compared with the state-of-the-art algorithms and has impressive efficiency and versatility using only the interactions between the two processes without incurring extra evaluations.Index Terms-Clustering, incremental machine learning, interacting processes, many-objective optimization, reference vector. * Equal contribution. This is an UPDATED preprint of https: //ieeexplore.ieee.org/document/

show abstract

Section: Motivationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Many-Objective Evolutionary Algorithm With Two Interacting Processes: Cascade Clustering and Reference Point Incremental Learning

Zhao

Sun

et al. 2019

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

show abstract

“…A performance analysis is done to evaluate the reliability of the solutions. Six multi-objective optimization algorithms are compared (1) ARMOEA [32]; (2) NSGA-II [33]; (3) SPEA2 [34]; (4) GrEA [35]; (5) RSEA [36] and VaEA [37].…”

Section: B the Whole Track Studymentioning

confidence: 99%

Pareto-Based Maintenance Decisions for Regional Railways With Uncertain Weld Conditions Using the Hilbert Spectrum of Axle Box Acceleration

Núñez

Jamshidi

Wang

2019

IEEE Trans. Ind. Inf.

View full text Add to dashboard Cite

Abstract1 -This paper presents a Pareto-based maintenance decision system for rail welds in a regional railway network. Weld health condition data are collected using a train in operation. A Hilbert spectrum-based approach is used for data processing to detect and assess the weld quality based on multiple registered dynamic responses in the axle box acceleration measurements. The assessment of the welds is stochastic in nature and variant over time, so a set of robust and predictive key performance indicators is defined to capture the weld degradation dynamics during a given maintenance period. Using a scenario-based approach, two objective functions are defined, performance and the number of weld replacements. Evolutionary multi-objective optimization is employed to optimize the objective functions so that the trade-offs between performance and cost support decision-making for railway network maintenance. The results of the proposed methodology show that the infrastructure manager can localize field inspections and maintenance efforts on the area with the most critical welds. To showcase the capability of the proposed methodology, measurements from a regional railway network in Transylvania, Romania are employed.

show abstract

“…Decomposition strategy, where a complex multi-objective problem is decomposed into several single-objective problems or several simpler multi-objective problems, such as NSGA-III [19] and MOEA/DM [20]. Performance indicator strategy, where measurement quality of solutions is used as criteria of selection, and two widely used approaches are hypervolume-based evolutionary algorithm (HypE) [21] and an indicator based multi-objective evolutionary algorithm with reference point (AR-MOEA) [22]. Consequently, such processing methods and improved strategies, which adopt to sensing characteristics of cameras on a road network, can be introduced into the coverage optimization issue of camera network.…”

Section: Related Workmentioning

confidence: 99%

A multi-objective scheduling optimization algorithm of a camera network for directional road network coverage

et al. 2018

View full text Add to dashboard Cite

Effective video monitoring systems require optimization of camera and road network coverage, to exploit fully the hardware and software solutions in smart city traffic applications. Monitoring requirements have grown increasingly diverse as scenes are becoming increasingly complex, thereby transforming the camera and road network coverage optimization issue into a nonlinear, high-dimension, and multi-objective problem. Previous research on this topic however, has focused on a single, specific optimization objective, which may result in invalid optimization results in actual applications. To extend this research, we propose a multi-objective scheduling optimization algorithm for a camera network that addresses the problem of directional road network coverage. In this solution, we incorporate an expanding parameter of main optical axes into particle swarm optimization algorithm. Our new strategy divides the range of main optical axes of all the cameras to control the scheduling number, achieving collaborative optimization of multiple objectives. In a simulated camera and road network, an experiment was designed for evaluating the effectiveness of the proposed method, comparing the distribution of optimization results with the global and local optimal solutions of the true value. A second experiment compared the distribution, performance and running time of the optimization results with different values of expanding parameter of main optical axes. A third experiment compared the performance of the optimization solutions with different values of camera parameters. The results showed that the proposed method can adapt to user application preference, and is effective and robust to schedule and allocate monitoring resources in different scenarios.

show abstract

An Indicator-Based Multiobjective Evolutionary Algorithm With Reference Point Adaptation for Better Versatility

Cited by 524 publications

References 62 publications

A Many-Objective Evolutionary Algorithm With Two Interacting Processes: Cascade Clustering and Reference Point Incremental Learning

A Many-Objective Evolutionary Algorithm With Two Interacting Processes: Cascade Clustering and Reference Point Incremental Learning

Pareto-Based Maintenance Decisions for Regional Railways With Uncertain Weld Conditions Using the Hilbert Spectrum of Axle Box Acceleration

A multi-objective scheduling optimization algorithm of a camera network for directional road network coverage

Contact Info

Product

Resources

About