Disruptive Innovations and Disruptive Assurance: Assuring Machine Learning and Autonomy

被引：12

作者：

Bloomfield, Robin ^{[1
,2
]}

Khlaaf, Heidy ^{[1
]}

Conmy, Philippa Ryan ^{[1
]}

Fletcher, Gareth ^{[1
]}

机构：

[1] Adelard LLP, London, England

[2] City Univ London, London, England

来源：

COMPUTER | 2019年 / 52卷 / 09期

关键词：

D O I：

10.1109/MC.2019.2914775

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Autonomous and machine learning-based systems are disruptive innovations and thus require a corresponding disruptive assurance strategy. We offer an overview of a framework based on claims, arguments, and evidence aimed at addressing these systems and use it to identify specific gaps, challenges, and potential solutions.

引用

页码：82 / 89

页数：8

共 14 条

[1]

[Anonymous], 2016, DRIVING SAFETY MANY

[2]

[Anonymous], 2017, Reluplex: An efficient smt solver for verifying deep neural networks

[3]

[Anonymous], 2018, 11281 PAS BRIT STAND

[4] Security-Informed Safety: Supporting Stakeholders with Codes of Practice [J].

Bloomfield, Robin ;

Bishop, Peter ;

Butler, Eoin ;

Stroud, Robert .

COMPUTER, 2018, 51 (08) :60-65

[5]

Gilmer Justin, 2018, Motivating the rules of the game for adversarial example research

[6]

Goodfellow I., 2015, P INT C LEARN REPR C

[7]

Huang X., 2016, ARXIV

[8] Robustness Testing of Autonomy Software [J].

Hutchison, Casidhe ;

Zizyte, Milda ;

Lanigan, Patrick E. ;

Guttendorf, David ;

Wagner, Michael ;

Le Goues, Claire ;

Koopman, Philip .

2018 IEEE/ACM 40TH INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING - SOFTWARE ENGINEERING IN PRACTICE TRACK (ICSE-SEIP 2018), 2018, :276-285

[9] Challenges in Autonomous Vehicle Testing and Validation [J].

Koopman, Philip ;

Wagner, Michael .

SAE INTERNATIONAL JOURNAL OF TRANSPORTATION SAFETY, 2016, 4 (01) :15-24

[10]

Papernot N., 2016, SCI SECURITY PRIVACY

← 1 2 →