Early-stage malware prediction using recurrent neural networks
- Submitting institution
-
Cardiff University / Prifysgol Caerdydd
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 96290451
- Type
- D - Journal article
- DOI
-
10.1016/j.cose.2018.05.010
- Title of journal
- Computers and Security
- Article number
- -
- First page
- 578
- Volume
- 77
- Issue
- -
- ISSN
- 0167-4048
- Open access status
- Compliant
- Month of publication
- May
- Year of publication
- 2018
- URL
-
http://dx.doi.org/10.1016/j.cose.2018.05.010
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
2
- Research group(s)
-
C - Cybersecurity, privacy and human centred computing
- Citation count
- 50
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper is the first to predict malicious behaviour using machine activity data when malware begins to execute on a desktop PC. Previous work only detected an attack after it was complete – our new algorithm predicts whether an executable is malicious or benign within 5 seconds at 94% accuracy, producing a step-change in capability for rapid defence against malware such as WannaCry – which cost the NHS £92m. Predictive algorithms based upon this work are now integrated within Airbus’s front-line Security Operation Centres – protecting Airbus’ 134,000 employees, confidential data, and key infrastructure across Europe.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -