Sarah Monazam Erfani

Closing the BIG-LID: An Effective Local Intrinsic Dimensionality Defense for Nonlinear Regression Poisoning

Sandamal Weerasinghe, Tamas Abraham, Tansu Alpcan, Sarah M Erfani, Christopher Leckie, Benjamin IP Rubinstein

Conference Proceedings  |  2021  |  Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence

Nonlinear regression, although widely used in engineering, financial and security applications for automated decision making, is k..

Are you with me? Measurement of learners' video-watching attention with eye tracking

N Srivastava, S Nawaz, J Newn, J Lodge, E Velloso, SM Erfani, D Gasevic, J Bailey

Conference Proceedings  |  2021  |  ACM International Conference Proceeding Series

Video has become an essential medium for learning. However, there are challenges when using traditional methods to measure how lea..

Defending Support Vector Machines Against Data Poisoning Attacks

Sandamal Weerasinghe, Tansu Alpcan, Sarah M Erfani, Christopher Leckie

Journal article  |  2021  |  IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY

Support Vector Machines (SVMs) are vulnerable to targeted training data manipulations such as poisoning attacks and label flips. B..

High Intrinsic Dimensionality Facilitates Adversarial Attack: Theoretical Evidence

Laurent Amsaleg, James Bailey, Amdeie Barbe, Sarah M Erfani, Teddy Furon, Michael E Houle, Milos Radovanovic, Vinh Nguyen Xuan

Journal article  |  2021  |  IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY

Machine learning systems are vulnerable to adversarial attack. By applying to the input object a small, carefully-designed perturb..

Segmented Pairwise Distance for Time Series with Large Discontinuities

Jiabo He, Sarah Erfani, Sudanthi Wijewickrema, Stephen O'Leary, Kotagiri Ramamohanarao

Conference Proceedings  |  2020  |  Proceedings of the International Joint Conference on Neural Networks

Time series with large discontinuities are common in many scenarios. However, existing distance-based algorithms (e.g., DTW and it..

Adversarial Reinforcement Learning under Partial Observability in Autonomous Computer Network Defence

Yi Han, David Hubczenko, Paul Montague, Olivier De Vel, Tamas Abraham, Benjamin IP Rubinstein, Christopher Leckie, Tansu Alpcan, Sarah Erfani

Conference Proceedings  |  2020  |  Proceedings of the International Joint Conference on Neural Networks

Recent studies have demonstrated that reinforcement learning (RL) agents are susceptible to adversarial manipulation, similar to v..

Discovery of contrast corridors from trajectory data in heterogeneous dynamic cellular networks

Li Li, Sarah Erfani, Chien Aun Chan, Christopher Leckie

Conference Proceedings  |  2020  |  Proceedings of the International Joint Conference on Neural Networks

In 5G networks, the deployment and management of mobile edge computing infrastructure is a major challenge for mobile operators. M..

Learning Non-Unique Segmentation with Reward-Penalty Dice Loss

Jiabo He, Sarah Erfani, Sudanthi Wijewickrema, Stephen O'Leary, Kotagiri Ramamohanarao

Conference Proceedings  |  2020  |  Proceedings of the International Joint Conference on Neural Networks

Semantic segmentation is one of the key problems in the field of computer vision, as it enables computer image understanding. Howe..

Heterogeneous task co-location in containerized cloud computing environments

Z Zhong, J He, MA Rodriguez, S Erfani, R Kotagiri, R Buyya

Conference Proceedings  |  2020  |  Proceedings - 2020 IEEE 23rd International Symposium on Real-Time Distributed Computing, ISORC 2020

Although cloud computing became a mainstream industrial computing paradigm, low resource utilization remains a common problem that..

LN-SNE: Log-Normal Distributed Stochastic Neighbor Embedding for Anomaly Detection

Zahra Ghafoori, Sarah M Erfani, James C Bezdek, Shanika Karunasekera, Christopher Leckie

Journal article  |  2020  |  IEEE Transactions on Knowledge and Data Engineering

We present a new unsupervised dimensionality reduction technique, called LN-SNE, for anomaly detection. LN-SNE generates a paramet..