Surrogate-Assisted Autoencoder-Embedded Evolutionary Optimization Algorithm to Solve High-Dimensional Expensive Problems

Cui, Meiji; Li, Li; Zhou, MengChu; Abusorrah, Abdullah

doi:10.1109/tevc.2021.3113923

Cited by 80 publications

(5 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• The effect of the dimension-reduction technique may be enhanced further by conducting the evolutionary search in a reduced search space, as ADSAPSO performs effectively. This tendency is consistent with recent observations on autoencoder-embedded approaches for singleobjective problems [63], [64].…”

Section: B Resultssupporting

confidence: 93%

Evolutionary multiobjective optimization assisted by scalarization function approximation for high-dimensional expensive problems

Horaguchi,

Nishihara,

Nakata

2024

Swarm and Evolutionary Computation

View full text Add to dashboard Cite

Section: B Resultssupporting

confidence: 93%

Evolutionary multiobjective optimization assisted by scalarization function approximation for high-dimensional expensive problems

Horaguchi,

Nishihara,

Nakata

2024

Swarm and Evolutionary Computation

View full text Add to dashboard Cite

“…We will also work towards the computational efficiency of the MIDE using parallel processing. Moreover, we will test the ability of the MIDE for solving large-scale problems [61] and problems having noisy environments [62]. Based on this performance, in the future, we will extend the MIDE for these types of problems.…”

Section: Discussionmentioning

confidence: 99%

Multiple Elite Individual Guided Piecewise Search-Based Differential Evolution

Gupta

Singh

et al. 2023

IEEE/CAA J. Autom. Sinica

View full text Add to dashboard Cite

The differential evolution (DE) algorithm relies mainly on mutation strategy and control parameters' selection. To take full advantage of top elite individuals in terms of fitness and success rates, a new mutation operator is proposed. The control parameters such as scale factor and crossover rate are tuned based on their success rates recorded over past evolutionary stages. The proposed DE variant, MIDE, performs the evolution in a piecewise manner, i.e., after every predefined evolutionary stages, MIDE adjusts its settings to enrich its diversity skills. The performance of the MIDE is validated on two different sets of benchmarks: CEC 2014 and CEC 2017 (special sessions & competitions on real-parameter single objective optimization) using different performance measures. In the end, MIDE is also applied to solve constrained engineering problems. The efficiency and effectiveness of the MIDE are further confirmed by a set of experiments.

show abstract

“…The DE algorithm has a number of operations that depend on the size of the population [37]. This approach is related to problem dimensionality, d, and population size.…”

Section: Dimension-dependent Population Sizingmentioning

confidence: 99%

Improving Performance of Differential Evolution Using Multi-Population Ensemble Concept

Bashir,

Abbas,

Mahmood

et al. 2023

Symmetry

View full text Add to dashboard Cite

Differential evolution (DE) stands out as a straightforward yet remarkably powerful evolutionary algorithm employed for real-world problem-solving purposes. In the DE algorithm, few parameters are used, and the population is evolved by applying various operations. It is difficult in evolutionary computation algorithms to maintain population diversity. The main issue is the sub-population of the DE algorithm that helps improve convergence speed and escape from the local optimum. Evolving sub-populations by maintaining diversity is an important issue in the literature that is considered in this research. A solution is proposed that uses sub-populations to promote greater diversity within the population and improve the algorithm performance. DE, heterogeneous distributed differential evolution (HDDE), multi-population ensemble differential evolution (MPEDE), and the proposed improved multi-population ensemble differential evolution (IMPEDE) are implemented using parameter settings; population sizes of 100 NP, 150 NP, and 200 NP; and dimensions of 10D, 30D, and 50D for performance comparison. Different combinations of mutations are used to generate the simulated results. The simulation results are generated using 1000, 3000, and 5000 iterations. Experimental outcomes show the superior results of the proposed IMPEDE over existing algorithms. The non-parametric significance Friedman test confirms that there is a significant difference in the performance of the proposed algorithm and other algorithms used in this study by considering a 0.05 level of significance using six benchmark functions.

show abstract

Surrogate-Assisted Autoencoder-Embedded Evolutionary Optimization Algorithm to Solve High-Dimensional Expensive Problems

Cited by 80 publications

References 69 publications

Evolutionary multiobjective optimization assisted by scalarization function approximation for high-dimensional expensive problems

Evolutionary multiobjective optimization assisted by scalarization function approximation for high-dimensional expensive problems

Multiple Elite Individual Guided Piecewise Search-Based Differential Evolution

Improving Performance of Differential Evolution Using Multi-Population Ensemble Concept

Contact Info

Product

Resources

About