深度学习应用于网络空间安全的现状、趋势与展望_图文

admin管理员组

文章数量:1530518

2023年12月29日发(作者：)

计算机研究与发展Journal of Computer Research and DevelopmentDOI：10. 7544/issnl000-1239. 2018. 20170649

55(6) : 1117-n)2, 2018深度学习应用于网络空间安全的现状、趋势与展望张玉清12董颖1柳彩云1雷柯楠12孙鸿宇12U中国科学院大学国家计算机网络入侵防范中心北京101408)2 (西安电子科技大学网络与信息安全学院西安710071)(zhangyq@nipc. org. cn)Situation, Trends and Prospects of Deep Learning Applied to Cyberspace SecurityZhangYuqing1，2 $ DongYing1 $ LiuCaiyun1 $ LeiKenan1，2 $ andSunHongyu1，21

(National Computer Netxvork Intrusion Protection Center $ University of Chinese Academy of Sciences $ Beijing 101408)2 {School of Cyber Engineering， Xidian University， Xi’an 7100711Abstract Recently， research on deep learning applied to cyberspace security has caused increasingacademic concern，and this survey analyzes the current research situation and trends of deep learning

applied to cyberspace security in terms of classification algorithms， feature extraction and learning

performance. Currently deep learning is mainly applied to malware detection

and this survey reveals the existing problems of these applications： feature selection，which could beachieved by extracting features from raw data； sel—adaptability，achieved by early-exit strategy to

update the model in real time； interpretability， achieved by influence functions to obtain the

correspondence between features and classification labels. Then，top 10 obstacles and opportunities in

deep learning research are summarized. Based on this， top 10 obstacles and opportunities of deep

learning applied to cyberspace security are at first proposed， which falls into three categories. The

first category is intrinsic vulnerabilities of deep learning to adversarial attacks and privacy-thett

attacks. The second category is sequence-model related problems，including program syntax analysis，

program code generation and long-term dependences in sequence modeling. The third category is

learning performance problems，including poor interpretability and traceability，poor sel--adaptability

and self-learning ability， false positives and data unbalance. Main obstacles and their opportunitiesamong the top 10 are analyzed，and we also point out that applications using classification models are

vulnerable to adversarial attacks and the most effective solution is adversarial training; collaborative

deep learning applica t i ons

in= vulnerable t o privacy- t het t at t acks，and prospective

student model. Finally， future research trends of deep learning applied to cyberspace security areKey

words

+ecuritydeep learning； cyberspace security； attacks and defenses； application security； network收稿日期！017-09-06 回日期：2018-01-17基金项目：国家重点研发计划项目（2016YFB0800703);国家自然科学基金项目（61572460,61272481);信息安全国家重点实验室的开放课题

(2017-ZD-01);国家发改委信息安全专项项目（（2012)1424)This work was supported by the National Key Research and Development Program of China (2016Natural Science Foundation of China (61572460，61272481)， the Open Program of the State Key Laboratory of InformationSecurity (2017-ZD-01)，and the Special Program on Information Security of the National Development and Reform Commission of

China ((2012)1424).

本文标签： 国家重点网络空间防范计算机网络