arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.08513cs.NE
收藏

基于双种群与均匀采样的多目标进化算法用于神经架构搜索

A Multi-objective Evolutionary Algorithm Based on Bi-population with Uniform Sampling for Neural Architecture Search

Nanjing University of Information Science and Technology(南京信息工程大学);China University of Mining and Technology(中国矿业大学);The Hong Kong Polytechnic University(香港理工大学);Macau University of Science and Technology(澳门科技大学);Qingdao University of Science and Technology(青岛科技大学)

Yu Xue, Pengcheng Jiang, Chenchen Zhu, Yong Zhang, Ran Cheng, Kaizhou Gao, Dunwei Gong

AI总结:

MOEA-BUS通过双种群与均匀采样方法,提升神经架构搜索中多目标优化的效率与性能。

摘要:

神经架构搜索(NAS)自动化神经网络设计,比手动方法更高效。然而,高效发现同时优化多个目标的高性能神经网络架构仍然是NAS中的重大挑战。现有方法往往存在种群多样性有限和搜索空间探索不足的问题,尤其是在极端复杂值区域。为了解决这些挑战,我们提出了MOEA-BUS,一种基于双种群与均匀采样的创新多目标进化算法,旨在同时优化准确性和网络复杂性。在MOEA-BUS中,提出了一种新的均匀采样方法来初始化种群,确保架构在目标空间中均匀分布。此外,为了增强探索能力,我们部署了双种群框架,其中两个种群协同进化,促进全面的搜索空间覆盖。在CIFAR-10和ImageNet上的实验表明,MOEA-BUS表现出色,实现了CIFAR-10上的98.39%的top-1准确率,以及ImageNet上的80.03%的准确率。值得注意的是,仅使用446M MAdds,MOEA-BUS在ImageNet上实现了78.28%的准确率。消融研究证实,均匀采样和双种群机制增强了种群多样性和性能。此外,在Kendall's tau系数方面,SVM相比其他三种常用机器学习模型至少提高了0.035,而均匀采样提供了约0.07的提升。

英文摘要:

Neural architecture search (NAS) automates neural network design, improving efficiency over manual approaches. However, efficiently discovering high-performance neural network architectures that simultaneously optimize multiple objectives remains a significant challenge in NAS. Existing methods often suffer from limited population diversity and inadequate exploration of the search space, particularly in regions with extreme complexity values. To address these challenges, we propose MOEA-BUS, an innovative multi-objective evolutionary algorithm based on bi-population with uniform sampling for neural architecture search, aimed at simultaneously optimizing both accuracy and network complexity. In MOEA-BUS, a novel uniform sampling method is proposed to initialize the population, ensuring that architectures are distributed uniformly across the objective space. Furthermore, to enhance exploration, we deploy a bi-population framework where two populations evolve synergistically, facilitating comprehensive search space coverage. Experiments on CIFAR-10 and ImageNet demonstrate MOEA-BUS's superiority, achieving top-1 accuracies of 98.39% on CIFAR-10, and 80.03% on ImageNet. Notably, it achieves 78.28% accuracy on ImageNet with only 446M MAdds. Ablation studies confirm that both uniform sampling and bi-population mechanisms enhance population diversity and performance. Additionally, in terms of the Kendall's tau coefficient, the SVM achieves an improvement of at least 0.035 compared to the other three commonly used machine learning models, and uniform sampling provided an enhancement of approximately 0.07.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑