Research projects in Information Technology

Anomaly detection methods address the need for automatic detection of unusual events with applications in cybersecurity. This project aims to address the efficacy of existing models when applied to real-life data. The goal is to generate new knowledge in the field of time series anomaly detection [1,2] through the invention of methods that effectively learn to generalise patterns of normal from real-life data.

The existing deep learning based time series classification (TSC) algorithms have some success in multivariate time series, their accuracy is not high when we apply them on brain EEG time series (65-70%). This is because there is a large variation between EEG data of different subjects, so a TSC model cannot generalise on unseen subjects well. In this research project we investigate self-supervised contrastive learning to encode the EEG data. This way we can better model the distribution of our EEG data before classifying it to different mental status. See a recent work here [1].

Anomaly detection is an important task in data mining. Traditionally most of the anomaly detection algorithms have been designed for ‘static’ datasets, in which all the observations are available at one time. In non-stationary environments on the other hand, the same algorithms cannot be applied as the underlying data distributions change constantly and the same models are not valid. Hence, we need to devise adaptive models that take into account the dynamically changing characteristics of environments and detect anomalies in ‘evolving’ data.

Time series are an ever growing form of data, generated by numerous types of sensors and automated processes. However, machine learning and deep learning methods for analysing time series are much less advanced than for other forms of data.

Our research is revolutionising the analysis of time series data. But it is early days, and many more impactful challenges are yet to be overcome.

This project is funded by the Australian Research Council and will be conducted as part of a large world-leading research team.