arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.07556cs.LGstat.ML
收藏

基于Weisfeiler-Lehman核贝叶斯优化的可解释神经结构搜索

Interpretable Neural Architecture Search via Bayesian Optimisation with Weisfeiler-Lehman Kernels

University of Oxford(牛津大学)

Binxin Ru, Xingchen Wan, Xiaowen Dong, Michael Osborne

摘要:

目前的神经结构搜索(NAS)策略只关注于找到一个单一的,好的,架构。它们几乎无法解释为什么特定的网络表现良好,或者如果我们想要进一步改进,我们应该如何修改架构。我们提出了一个贝叶斯优化(BO)的NAS方法,结合Weisfeiler-Lehman图形内核与高斯过程代理。我们的方法优化的架构,在一个高度的数据效率的方式:它是能够捕获的拓扑结构的架构,并可扩展到大型图形,从而使高维和图形的搜索空间服从BO。更重要的是,我们的方法通过发现有用的网络特征及其对网络性能的相应影响来提供可解释性。事实上,我们的经验表明,我们的代理模型是能够识别有用的图案,可以指导新架构的生成。我们最后表明,我们的方法优于现有的NAS方法,以实现最先进的封闭和开放域搜索空间。

英文摘要:

Current neural architecture search (NAS) strategies focus only on finding a single, good, architecture. They offer little insight into why a specific network is performing well, or how we should modify the architecture if we want further improvements. We propose a Bayesian optimisation (BO) approach for NAS that combines the Weisfeiler-Lehman graph kernel with a Gaussian process surrogate. Our method optimises the architecture in a highly data-efficient manner: it is capable of capturing the topological structures of the architectures and is scalable to large graphs, thus making the high-dimensional and graph-like search spaces amenable to BO. More importantly, our method affords interpretability by discovering useful network features and their corresponding impact on the network performance. Indeed, we demonstrate empirically that our surrogate model is capable of identifying useful motifs which can guide the generation of new architectures. We finally show that our method outperforms existing NAS approaches to achieve the state of the art on both closed- and open-domain search spaces.

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。

↑