ModuleNet: Knowledge-Inherited Neural Architecture Search

doi:10.1109/TCYB.2021.3078573

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 深度强化学习

	ModuleNet: Knowledge-Inherited Neural Architecture Search
	Chen, Yaran1,2 ; Gao, Ruiyuan 3; Liu, Fenggang 4; Zhao, Dongbin1,2
发表期刊	IEEE TRANSACTIONS ON CYBERNETICS
ISSN	2168-2267
	2021-06-04
页码	11
通讯作者	Zhao, Dongbin(dongbin.zhao@ia.ac.cn)
摘要	Although neural the architecture search (NAS) can bring improvement to deep models, it always neglects precious knowledge of existing models. The computation and time costing property in NAS also means that we should not start from scratch to search, but make every attempt to reuse the existing knowledge. In this article, we discuss what kind of knowledge in a model can and should be used for a new architecture design. Then, we propose a new NAS algorithm, namely, ModuleNet, which can fully inherit knowledge from the existing convolutional neural networks. To make full use of the existing models, we decompose existing models into different modules, which also keep their weights, consisting of a knowledge base. Then, we sample and search for a new architecture according to the knowledge base. Unlike previous search algorithms, and benefiting from inherited knowledge, our method is able to directly search for architectures in the macrospace by the NSGA-II algorithm without tuning parameters in these modules. Experiments show that our strategy can efficiently evaluate the performance of a new architecture even without tuning weights in convolutional layers. With the help of knowledge we inherited, our search results can always achieve better performance on various datasets (CIFAR10, CIFAR100, and ImageNet) over original architectures.
关键词	Computer architecture Task analysis Knowledge based systems Microprocessors Statistics Sociology Computational modeling Evaluation algorithm knowledge inherited neural architecture search (NAS)
DOI	10.1109/TCYB.2021.3078573
关键词[WOS]	GENETIC ALGORITHM ; MODEL
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China (NSFC)[62006226] ; Youth Research Fund of the State Key Laboratory of Management and Control for Complex Systems[20190213] ; Huawei Technologies Company Ltd.[FA2018111061SOW12]
项目资助者	National Natural Science Foundation of China (NSFC) ; Youth Research Fund of the State Key Laboratory of Management and Control for Complex Systems ; Huawei Technologies Company Ltd.
WOS研究方向	Automation & Control Systems ; Computer Science
WOS类目	Automation & Control Systems ; Computer Science, Artificial Intelligence ; Computer Science, Cybernetics
WOS记录号	WOS:000732096300001
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
七大方向——子方向分类	强化与进化学习
引用统计	被引频次：11[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/46906
专题	多模态人工智能系统全国重点实验室_深度强化学习
通讯作者	Zhao, Dongbin
作者单位	1.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Coll Artificial Intelligence, Beijing 100049, Peoples R China 3.Chinese Univ Hong Kong, Dept Comp Sci & Engn, Hong Kong, Peoples R China 4.Beijing Inst Technol, Coll Automat, Beijing 100811, Peoples R China
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Chen, Yaran,Gao, Ruiyuan,Liu, Fenggang,et al. ModuleNet: Knowledge-Inherited Neural Architecture Search[J]. IEEE TRANSACTIONS ON CYBERNETICS,2021:11.
APA	Chen, Yaran,Gao, Ruiyuan,Liu, Fenggang,&Zhao, Dongbin.(2021).ModuleNet: Knowledge-Inherited Neural Architecture Search.IEEE TRANSACTIONS ON CYBERNETICS,11.
MLA	Chen, Yaran,et al."ModuleNet: Knowledge-Inherited Neural Architecture Search".IEEE TRANSACTIONS ON CYBERNETICS (2021):11.