ML-CSS@ICL'22

Details

Location: Dyson School of Design Engineering – Lecture theatre G01B, SW7 2AZ. (Enter through Fusion Cafe)

Date: Tuesday, May 31, 2022.

Time: 11:00-17:00

Lightning talks (~15 minutes), keynote preseantations, and networking!

Coffee/tea and snacks (vegan/gf options available) provided by the EEE department.

We invite all researchers within Imperial to attend.
(e.g. postgraduate students, researchers, and professors - and invited guests from this audience)

For further information please contact us on f[dot]alotaibi21[at]imperial[dot]ac[dot]uk.

Keynote Speakers

Trustworthy Machine Learning... for Systems Security

Presented by:

Certified Federated Adversarial Training

Presented by:

Schedule

This symposium is composed of INFORMAL lightning talks by researchers at Imperial who work in the intersection of machine learning and cyber security. The talks should be about a current applied ML&Security research (COMPLETE OR NOT COMPLETE!!), which will be followed by brief Q&A.

The goal of this event is to connect those who work in this area; we look forward to your active participation.

Time	Speaker	Talk Title
11:00	Lorenzo Cavallaro	Trustworthy Machine Learning... for Systems Security
11:30	Eman Maali	Online Anomaly detection for IoT Networks
11:45	Tony Ng	NinjaDesc: Content-Concealing Visual Descriptors via Adversarial Learning
12:00	---	LUNCH
13:00	Giulio Zizzo	Certified Federated Adversarial Training
13:30	Fahad Alotaibi	Concept Drift in Network Intrusion Detection Systems
13:45	Shubham Jain	Adversarial Detection Avoidance Attacks: Evaluating the robustness of perceptual hashing-based client-side scanning
14:00	Wangkun Xu	Blending Data and Physics Against False Data Injection Attack: An Event-Triggered Moving Target Defence Approach
14:15	Francesco Sanna Passino	Unsupervised attack pattern detection in cyber-security using Bayesian topic modelling
14:30	Almuthanna Alageel	Hawk-Eye: Holistic Detection of APT Command and Control Domains
14:45	Tom Davies	Topological data analysis for anomaly detection in host-based logs
15:00	Hazim Hanif	Neural representation learning for software vulnerability detection
15:15	Moh Malekzadeh	Honest-but-Curious Nets: Sensitive Attributes of Private Inputs Can Be Secretly Coded into the Classifiers' Outputs
15:30	Kate Highnam	The Cake is a Lie: Self-Adaptation for Black-Box Models Problem Scenario
15:45	Myles Foley	Hacking websites with Reinforcement Learning: an XSS story
16:00	---	SNACK BREAK
16:15	Networking Session	Practices for Cyber Security and Privacy
17:00	Closing Remarks

If you are a speaker and the assigned time does not work for you, please contact Kate Highnam (kwh19[at]ic[dot]ac[dot]uk).

Organisers

Kate Highnam

Ph.D. Student in Intelligent Systems and Networks within EEE

Kate is a Ph.D. Student under the joint supervision of Professor Nicholas R. Jennings CB, FREng, and Dr. Sergio Maffeis. Her professional experience in machine learning and cyber security motivates her current research into domain adaptation in intrusion detection with real world applications. She is also an Enrichment student with The Alan Turing Institute in London.

Eman Maali

Ph.D. Student in Machine Learning for IoT security at the DE

Emaan is a Ph.D. candidate in the Dyson School of Design Engineering at The Faculty of Engineering. She is working under the supervision of Dr. Hamed Haddadi and Dr.David Boyle. Her research in the field of machine learning for a secured IoT environment. In 2017, She completed her MSc in Electromagnetic Sensor Networks at the University of Birmingham. The focus of the Masters was on electromagnetic, antennas, propagation, computer communications networks and RF and microwave engineering. Moreover, She completed her BA in Computer Systems Engineering from Birzeit University in Palestine.

Myles Foley

Ph.D. Student in Renforcement Learning for Cyber Security at the DoC.

Myles is a PhD student at Imperial College London under the supervision of Dr. Maffeis. He received his MEng from University College London in Electronic Engineering with Computer Science, earning the ‘Outstanding MEng Graduating Student’ prize. Myles’ research is focused at novel - and exciting - ways of applying reinforcement learning to problems in cyber security.

Fahad Alotaibi

Ph.D. Student in Learning-based Security Applications Security and Robustness at the DoC.

Fahad is a PhD student at Imperial College London under the supervision of Dr. Maffeis. He received his MSc from The University of York (UK) in Cyber Security, and his BCs from Shaqra University (KSA) in Computer Science. Fahad’ research is focused on robusting deep learning-based security applications againsts concept drift and poisoning attacks. Fahad is also interested in other areas such as digital forensics and ransomware prevention.