.png){kind=icon}
返回
Dynamic matrix-based evolutionary algorithm for large-scale sparse multiobjective optimization problems
.png){kind=icon}
PRE
AI
DOI：10.1007/s12293-023-00394-z.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Multiobjective optimization problems exist widely in practical applications involving large-scale decision variables and sparse Pareto optimal solutions. However, in terms of sparse detection, the existing sparse Large-scale MOEAs ignore global interaction characteristics in the evolutionary stage, which easily leads to the loss of key variables and inhibits convergence. Based on that, this paper proposes the use of a matrix to evaluate the decision variable correlations and perform adaptive updates, which achieve the highest accuracy of sparse relationship mining. Specifically, the control variable method and Pareto dominance relations generate an original matrix in the initial stage. Then, with reference to the mask distribution of the contemporary nondominated solutions, the matrix is updated in the evolutionary phase by means of the nondominated ordering of a parent and the corresponding offspring for rewards and penalties. Furthermore, a new genetic operator is proposed to ensure the sparsity of the generated solutions. According to the experimental results of eight benchmark problems, the algorithm outperforms existing evolutionary algorithms in solving sparse LSMOPs.
Keyword:
Evolutionary algorithm
Large-scale multiobjective optimization
Genetic operator
Sparse pareto optimal solutions
期刊
IF:
2.3
论文数:
453
被引数:
718
机构
引用论文
Solving Large-Scale Multiobjective Optimization Problems With Sparse Optimal Solutions via Unsupervised Neural Networks基于无监督神经网络求解具有稀疏最优解的大规模多目标优化问题
OAAI
OAAI分享收藏PlatEMO: A MATLAB Platform for Evolutionary Multi-Objective OptimizationPlatEMO: 用于进化多目标优化的MATLAB平台OAAI
OAAI分享收藏Solving large-scale multiobjective optimization via the probabilistic prediction model基于概率预测模型的大规模多目标优化求解.png)
MEMETIC COMPUTING
MEMETIC COMPUTINGIF2.3
PREAI
分享收藏分享收藏
查看更多内容