A multi-objective bilevel optimisation evolutionary algorithm with dual populations lower-level search

Wang, Weizhong; Liu, Hailin; Shi, Hongcan

doi:10.1080/09540091.2022.2077312

Cited by 8 publications

(2 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method applied an interaction matrix to group variables that are optimized in a cooperative manner. Wang et al [20] designed a double-population LL search strategy, where the first population maintains the convergence and diversity in the LL objective space, and the second population remains the non-dominance at both the UL and LL. Furthermore, Deb et al [6] introduced a simple and efficient bi-level multi-objective optimization algorithm to minimize the deviation in the expected outcome of the UL due to the independent decision-making process for LL.…”

Section: B Bl-mopsmentioning

confidence: 99%

Bi-Level Multiobjective Evolutionary Learning: A Case Study on Multitask Graph Neural Topology Search

Wang

Jiao

Zhao

et al. 2024

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

The construction of machine learning models involves many bi-level multi-objective optimization problems (BL-MOPs), where upper level (UL) candidate solutions must be evaluated via training weights of a model in the lower level (LL). Due to the Pareto optimality of sub-problems and the complex dependency across UL solutions and LL weights, an UL solution is feasible if and only if the LL weight is Pareto optimal. It is computationally expensive to determine which LL Pareto weight in the LL Pareto weight set is the most appropriate for each UL solution. This paper proposes a bi-level multi-objective learning framework (BLMOL), coupling the above decisionmaking process with the optimization process of the UL-MOP by introducing LL preference r. Specifically, the UL variable and r are simultaneously searched to minimize multiple UL objectives by evolutionary multi-objective algorithms. The LL weight with respect to r is trained to minimize multiple LL objectives via gradient-based preference multi-objective algorithms. In addition, the preference surrogate model is constructed to replace the expensive evaluation process of the UL-MOP. We consider a novel case study on multi-task graph neural topology search. It aims to find a set of Pareto topologies and their Pareto weights, representing different trade-offs across tasks at UL and LL, respectively. The found graph neural network is employed to solve multiple tasks simultaneously, including graph classification, node classification, and link prediction. Experimental results demonstrate that BLMOL can outperform some state-of-the-art algorithms and generate well-representative UL solutions and LL weights.

show abstract

Section: B Bl-mopsmentioning

confidence: 99%

Bi-Level Multiobjective Evolutionary Learning: A Case Study on Multitask Graph Neural Topology Search

Wang

Jiao

Zhao

et al. 2024

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

show abstract

“…In order to evaluate the performance of the algorithms, two algorithms were chosen for comparison: BLMOCC [20] , MOBEA-DPL [27] .…”

Section: 2compared Algorithms and Parameter Settingsmentioning

confidence: 99%

A cluster-based co-evolutionary optimization method for bilevel multi-objective optimization

Hu,

Li,

Huang

et al. 2024

Journal of Smart Environments and Green Computing

View full text Add to dashboard Cite

The research motivation of multi-objective bilevel optimization mainly stems from the need to solve practical problems and improve decision-making efficiency. On the one hand, bilevel optimization helps to solve the complexity and uncertainty in real life, thereby improving decision-making efficiency and robustness. On the other hand, by promoting the development and application of AI technology, bilevel optimization also provides support for sustainable development. Although the application of bilevel optimization has proven to be beneficial in addressing various real-life problems. However, recent studies indicate that achieving both high speed and high-quality optimization through existing algorithms remains challenging. This difficulty arises due to the NP-hard nature of the bilevel optimization problem. The nested structure method, commonly used to tackle this problem, involves each upper level solution independently performing the lower level optimization task. This approach significantly increases the number of evaluations for the lower level. To address this issue, our proposed method leverages the similarity in lower level optimization to group upper level solutions, enabling co-evolution of lower level solutions within the same group. Consequently, this approach substantially reduces the number of evaluations required for lower level solutions. Additionally, our method pairs parents and offspring, the optimized lower level solutions of the parents are utilized to optimize the lower level solutions of the offspring. This approach accelerates the optimization process for the lower level. To validate the effectiveness of our algorithm, we have applied it to a suite of test problems, demonstrating satisfactory performance.

show abstract

A Cluster-Based Dynamic Grouping Population Replication Strategy for Bilevel Multi-objective Optimization

et al. 2023

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

A multi-objective bilevel optimisation evolutionary algorithm with dual populations lower-level search

Cited by 8 publications

References 32 publications

Bi-Level Multiobjective Evolutionary Learning: A Case Study on Multitask Graph Neural Topology Search

Bi-Level Multiobjective Evolutionary Learning: A Case Study on Multitask Graph Neural Topology Search

A cluster-based co-evolutionary optimization method for bilevel multi-objective optimization

A Cluster-Based Dynamic Grouping Population Replication Strategy for Bilevel Multi-objective Optimization

Contact Info

Product

Resources

About