Solving High-Dimensional Multi-Objective Optimization Problems with Low Effective Dimensions

Authors: Hong Qian, Yang Yu

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results indicate that Re MO is effective for optimizing the high-dimensional MO functions with low effective dimensions, and is even effective for the high-dimensional MO functions where all dimensions are effective but most only have a small and bounded effect on the function value.
Researcher Affiliation Academia Hong Qian, Yang Yu National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, 210023, China Collaborative Innovation Center of Novel Software Technology and Industrialization, Nanjing, 210023, China {qianh,yuy}@lamda.nju.edu.cn
Pseudocode Yes Algorithm 1 Multi-Objective Optimization via Random Embedding (Re MO)
Open Source Code No The paper does not provide an explicit statement or link for the open-sourcing of the code for the described methodology.
Open Datasets Yes We first verify the effectiveness of Re MO empirically on three high-dimensional bi-objective (i.e., m = 2) optimization testing functions ZDT10, ZDT20 and ZDT30 with low M-effective dimensions. They are constructed based on the first three bi-objective functions ZDT1, ZDT2 and ZDT3 introduced in (Zitzler, Deb, and Thiele 2000).
Dataset Splits No The paper uses synthetic test functions (ZDT series) for which standard train/validation/test splits are not applicable in the same way as for empirical datasets. No explicit details on data partitioning or validation splits are provided.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory amounts) used for running its experiments.
Software Dependencies No The paper mentions several algorithms like NSGA-II and MOEA/D, stating 'The implementations of them are both by their authors,' but it does not provide specific version numbers for these or any other software dependencies.
Experiment Setup Yes We set the function evaluation budget n = 0.3D = 3000 and the upper bound of M-effective dimension ϑ = 50 > ϑe = 30. We set the function evaluation budget n = 0.3D = 3000 and set the reference point as (1, 4) to calculate the hyper-volume indicator (the larger the better).