Fabio Di Troia

Dr. Fabio Di Troia

orcid

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 April 2020 in Expert Systems with Applications
Verified
Journal Article
Closed access
MWMayuri Wadkar
Mayuri Wadkar
FTFabio Di Troia
Fabio Di Troia
MSMark Stamp
Mark Stamp

1 January 2018
Verified
Conference Paper
Open access
SYSravani Yajamanam
Sravani Yajamanam
VSVikash Raja Samuel Selvin
Vikash Raja Samuel Selvin
FTFabio Di Troia
Fabio Di Troia
MSMark Stamp
Mark Stamp

1 January 2017
Verified
Conference Paper
Open access
AKAnkita Kapratwar
Ankita Kapratwar
FTFabio Di Troia
Fabio Di Troia
MSMark Stamp
Mark Stamp

27 January 2016 in Journal of Computer Virology and Hacking Techniques
Verified
Journal Article
Closed access
PDF file
SPSwathi Pai
Swathi Pai
FTFabio Di Troia
Fabio Di Troia
CVCorrado Aaron Visaggio
Corrado Aaron Visaggio
TAThomas Austin
Thomas Austin
MSMark Stamp
Mark Stamp

18 March 2024 in Journal of Computer Virology and Hacking Techniques
Verified
Journal Article
Open access
PDF file
MKMatouš Kozák
Matouš Kozák
MJMartin Jureček
Martin Jureček
MSMark Stamp
Mark Stamp
FTFabio Di Troia
Fabio Di Troia
Because of its world-class results, machine learning (ML) is becoming increasingly popular as a go-to solution for many tasks. As a result, antivirus developers are incorporating ML models into their toolchains. While these models improve malware detection capabilities, they also carry the disadvantage of being susceptible to adversarial attacks. Although this vulnerability has been demonstrated for many models in white-box settings, a black-box scenario is more applicable in practice for the domain of malware detection. We present a method of creating adversarial malware examples using reinforcement learning algorithms. The reinforcement learning agents utilize a set of functionality-preserving modifications, thus creating valid adversarial examples. Using the proximal policy optimization (PPO) algorithm, we achieved an evasion rate of 53.84% against the gradient-boosted decision tree (GBDT) detector. The PPO agent previously trained against the GBDT classifier scored an evasion rate of 11.41% against the neural network-based classifier MalConv and an average evasion rate of 2.31% against top antivirus programs. Furthermore, we discovered that random application of our functionality-preserving portable executable modifications successfully evades leading antivirus engines, with an average evasion rate of 11.65%. These findings indicate that ML-based models used in malware detection systems are sensitive to adversarial attacks and that better safeguards need to be taken to protect these systems.

28 August 2018 in Journal of Computer Virology and Hacking Techniques
Verified
Journal Article
Closed access
PDF file
ARAditya Raghavan
Aditya Raghavan
FTFabio Di Troia
Fabio Di Troia
MSMark Stamp
Mark Stamp