An Evolutionary Many-Objective Optimization Algorithm Based on Dominance and Decomposition

By Ke Li, Kalyanmoy Deb, Qingfu Zhang, Sam Kwong. IEEE Transactions on Evolutionary Computation, 2015

1,080 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics

View Paper: https://www.semanticscholar.org/paper/63e021e35c1ca0a038f81e5540fbba2cf1e7862b

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

1. Yaochu Jin
2. Xing-yi Zhang
3. Ye Tian
4. Ran Cheng
5. Hisao Ishibuchi
6. Markus Olhofer
7. Kay Chen Tan
8. Handing Wang

In Computational Theory and Mathematics

1. Yaochu Jin
2. Xing-yi Zhang
3. Ye Tian
4. Ran Cheng
5. Markus Olhofer
6. Hisao Ishibuchi
7. Kay Chen Tan
8. Handing Wang

Citations in the corpus

• Test Problems for Large-Scale Multiobjective and Many-Objective Optimization (2017, neutral): “As introduced in Section V-B, the fitness landscapes of a test problem can be defined by relating g I and g II to different basic single-objective functions ( η 1 – η 6 ) according to (17) and (18).”
• A benchmark test suite for evolutionary many-objective optimization (2017, methodology): “…the decomposition-based algorithms such as the NSGA-III [5], and the evolutionary algorithms based on both dominance and decomposition (MOEA/DD) [6], and the reference vector-guided evolutionary algorithm (RVEA) [7]; the performance indicator-based algorithms such as the fast hypervolume-based…”
• A Strength Pareto Evolutionary Algorithm Based on Reference Direction for Multiobjective and Many-Objective Optimization (2017, methodology): “To avoid such a situation, a two-layer (boundary and inside layers) approach for objectives over seven was proposed in [13] and [31], which uses the systematic approach to generate two reference direction sets: one set on the boundary layer and the other on the inside layer.”
• Two-Archive Evolutionary Algorithm for Constrained Multiobjective Optimization (2017, methodology): “For C-DTLZ benchmark problems, we use the method developed in [27] to fulfill this purpose.”
• Deep Reinforcement Learning for Multiobjective Optimization (2019, background): “Decomposition Strategy: Decomposition, as a simple yet efficient way to design the multiobjective optimization algorithms, has fostered a number of researches in the community, for example, cellular-based MOGA [26], MOEA/D, MOEA/DD [27], DMOEA-ε C [28], and NSGA-III [29].”
• A Surrogate-Assisted Reference Vector Guided Evolutionary Algorithm for Computationally Expensive Many-Objective Optimization (2018, methodology): “In [4, 48], 100000 reference points were used irrespective of the number of objectives, while in [52], only 500 reference points”
• A Survey of Multiobjective Evolutionary Algorithms Based on Decomposition (2017, methodology): “4) The studies MOEA/D-M2M [70], NSGA-III [40], MOEA/DD [39], and RVEA [41] provide new direction”
• A Reference Vector Guided Evolutionary Algorithm for Many-Objective Optimization (2016, methodology): “Another example is a recently proposed evolutionary manyobjective optimization algorithm based on both dominance and decomposition (MOEA/DD) [41], the motivation of which is to exploit the merits offered by both dominance and decomposition-based approaches.”
• An easy-to-use real-world multi-objective optimization problem suite (2020, result): “While MOEA/D-PBI showed a good performance on many-objective problems in previous studies [63, 7], it does not work well in our study.”
• A Survey on ensemble learning under the era of deep learning (2021, methodology): “Employing a multi-objective evolutionary algorithm (MOEA/ DD) (Li et al. 2015), multiple base deep learners were generated based on three deep learning architectures, including DBN (Hinton et al. 2006), ResNet (He et al. 2016) and deep auto-encoder (DAE) (Jia et al. 2016).”
• A Decision Variable Clustering-Based Evolutionary Algorithm for Large-Scale Many-Objective Optimization (2018, methodology): “NSGA-II (NSGA-III) [37], dominance and decomposition based MOEA [38], and the recently proposed reference vector”
• Performance of Decomposition-Based Many-Objective Algorithms Strongly Depends on Pareto Front Shapes (2017, methodology): “In MOEA/DD [31], a (μ + 1)ES-style model is used as in the original MOEA/D.”
• A Vector Angle-Based Evolutionary Algorithm for Unconstrained Many-Objective Optimization (2017, methodology): “For SBX operator, its distribution index is ηc = 30, and the distribution index of mutation operator is ηm = 20 [8], [11].”
• A Strengthened Dominance Relation Considering Convergence and Diversity for Evolutionary Many-Objective Optimization (2018, methodology): “The third idea is to combine Pareto dominance with decomposition-based algorithms, such as MOEA/DD [18] and BCE-MOEA/D [19].”
• Balancing Objective Optimization and Constraint Satisfaction in Constrained Evolutionary Multiobjective Optimization (2021, background): “…ones, the algorithm in [25] makes no difference when the constraint violations of two solutions are similar, the algorithm in [26] treats two infeasible solutions as nondominated if the angle between them is large, and the algorithm in [27] preserves the infeasible solutions in isolated regions.”