期刊:
E3S Web of Conferences,2021年293:03025-null ISSN:2267-1242
通讯作者:
Wang, Z.
作者机构:
[Zhang K.; Lv J.; Wang Z.; Yuan Z.; Li X.; Yi H.] School of Architecture, University of South China, Hengyang, 421001, China;[Bedra K.B.] School of Architecture and Art, Central South University, Changsha, 410083, China
通讯机构:
[Wang, Z.] S;School of Architecture, China
会议名称:
3rd Global Conference on Ecological Environment and Civil Engineering, GCEECE 2021
会议论文集名称:
IEEE Congress on Evolutionary Computation
摘要:
Studies have shown that deep neural networks (DNNs) are susceptible to adversarial attacks, which can cause misclassification. The adversarial attack problem can be regarded as an optimization problem, then the genetic algorithm (GA) that is problem-independent can naturally be designed to solve the optimization problem to generate effective adversarial examples. Considering the dimensionality curse in the image processing field, traditional genetic algorithms in high-dimensional problems often fall into local optima. Therefore, we propose a GA with multiple fitness functions (MF-GA). Specifically, we divide the evolution process into three stages, i.e., exploration stage, exploitation stage, and stable stage. Besides, different fitness functions are used for different stages, which could help the GA to jump away from the local optimum. Experiments are conducted on three datasets, and four classic algorithms as well as the basic GA are adopted for comparisons. Experimental results demonstrate that MF-GA is an effective black-box attack method. Furthermore, although MF-GA is a black-box attack method, experimental results demonstrate the performance of MF-GA under the black-box environments is competitive when comparing to four classic algorithms under the white-box attack environments. This shows that evolutionary algorithms have great potential in adversarial attacks.
作者机构:
Sun Yat Sen Univ, Guangzhou, Peoples R China.;Sun Yat Sen Univ, Affiliated Hosp 3, Guangzhou, Peoples R China.;Chinese Univ Hong Kong, Sch Life Sci, Hong Kong, Peoples R China.;Foshan First Municipal Peoples Hosp, Foshan, Peoples R China.;Univ South China, Affiliated Hosp 1, Hengyang, Peoples R China.
会议名称:
Annual Meeting of the American-Society-of-Clinical-Oncology (ASCO)
作者机构:
[Chen Hai-ting; Wang Zhi-yuan] Univ South China, Sch Architecture, Hengyang 421001, Hunan, Peoples R China.;[Wang Zhi-yuan] Cent South Univ, Sch Architecture & Art, Changsha 41007, Hunan, Peoples R China.;[Ou Yang-mei] Hunan Inst Traff Engn, Hengyang 42100, Hunan, Peoples R China.
会议名称:
4th International Workshop on Renewable Energy and Development (IWRED)
会议时间:
APR 24-26, 2020
会议地点:
ELECTR NETWORK
会议主办单位:
[Wang Zhi-yuan;Chen Hai-ting] Univ South China, Sch Architecture, Hengyang 421001, Hunan, Peoples R China.^[Wang Zhi-yuan] Cent South Univ, Sch Architecture & Art, Changsha 41007, Hunan, Peoples R China.^[Ou Yang-mei] Hunan Inst Traff Engn, Hengyang 42100, Hunan, Peoples R China.
会议论文集名称:
IOP Conference Series-Earth and Environmental Science
摘要:
Regional urban connection is mainly reflected in the connection between transportation and economy.Using accessibility and improved gravity model, this paper studies the traffic and economic relations and spatial pattern analysis among districts and counties in Hengyang based on the shortest travel time data and socio-economic data of 12 districts and counties in Hengyang, China.The results show that: firstly, in terms of traffic accessibility, there are 8 districts and counties in Hengyang that have higher accessibility than the average level of the city, and 4 districts and counties that have lower accessibility than the average level. Secondly, the accessibility of districts and counties in Hengyang generally presents an irregular circular distribution pattern, forming a core edge distribution feature with five districts in Hengyang as the core and gradually decreasing outward enclosure.Thirdly, from the perspective of economic connection, the 12 districts and counties in Hengyang are divided into five levels. The first level is the connection between Shigu District and Yanfeng District in the central city. The second level connection intensity is between Shigu District and Zhengxiang District in the central city and between Zhengxiang District and Hengyang County. The intensity of the three-level connection is the connection between the counties near the central urban area and the districts in the central urban area, as well as the connection between the counties near the central urban area and the neighboring counties. The fourth level connection refers to the connection between districts and counties far away from the central urban area and counties near the central urban area. The fifth level connection refers to the connection between districts and counties far away from the central urban area. Last but not least, the economic connection of Hengyang District and county forms the center-peripheral radiation pattern which diffuses from the area with the core of Shigu District, steamed Xiang District, Yanfeng District and Zhuhui District as the core, which presents the pattern of "one core and two centers" in space.