Please feel free to email me with questions or requests concerning the publications. I will try and help as much as I can.
Online anomaly rate parameter tracking for anomaly detection in wireless sensor networks PDF
Colin O'Reilly, Alex Gluhak, Muhammad Ali Imran. Proceedings of 9th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks (SECON 2012), pages 191-199, 2012.
We examine a method for the automatic determination of parameters for a varient of the one-class support vector machine.
Online anomaly detection with an incremental centred kernel hypersphere PDF
Colin O'Reilly, Alex Gluhak, Muhammad Ali Imran. Proceedings of 2013 IEEE International Workshop on Machine Learning for Signal Processing (MLSP), 2013.
We examine the performance of using a kernel space to perform anomaly detection.
Adaptive anomaly detection with kernel eigenspace splitting and merging PDF
Colin O'Reilly, Alex Gluhak, Muhammad Ali Imran. IEEE Transactions on Knowledge and Data Engineering, 27(1), pages 3-16, 2015.
An incremental version of kernel PCA is derived. This is then used to perform anomaly detection in a non-stationary environment where regular model updates are required. The performance is compared with other incremental kernel PCA techniques.
Distributed Anomaly Detection using Minimum Volume Elliptical Principal Component Analysis PDF
Colin O'Reilly, Alex Gluhak, Muhammad Ali Imran. IEEE Transactions on Knowledge and Data Engineering 28, no. 9 (2016): 2320-2333.
Univariate and Multivariate Time Series Manifold Learning
Colin O'Reilly, Klaus Moessner, Michele Nati. Knowledge-Based Systems, 2017. PDF