Large amounts of data are created daily through web query logs, blogs, social media posts as well as information captured by satellites, sensors or medical devices.
Data science research in CDU focuses on data mining, data management, machine learning, data analysis and information visualisation. The purpose of the research group is to create systems and algorithms to extract useful information, finding correlations and predict trends from large databases for various applications and visualisation.
Research areas include:
- Managing, sorting and extracting useful information from vast amounts of data in the field of medical sciences, cyber security, bioinformatics and engineering.
- Managing and integrating different forms of data such as text, audio, video or images.
- Finding ways to increase the processing speed through enhanced algorithms or integration of high performing computing platforms.
- Designing more user-friendly platforms to improve the communication systems and feedback between the machines, algorithms and people.
Parameterized Complexity as a methodology is a specific focus of this research group. It's a recent branch of computational complexity theory that provides a framework for a refined analysis of hard algorithmic problems.
The big, important problems facing planet Earth have a structure with “secondary” measurements (parameters), apart from the primary measurement of overall input size, that significantly affects computational problem complexity.
The central notion of fixed parameter tractability (FPT) is a generalisation of polynomial-time based on confining any non-polynomial (typically exponential) complexity costs to a function only of these secondary measurements. Parameterised algorithms have strong connections to heuristics for NP-hard problems, and the multivariate approach allows more realistic modelling of real-world input distributions.
Potential areas of PhD and Master by Research topics include:
- Foundational multivariate algorithmics; including, for example, research into interconnections with heuristics and approximations.
- Applications in analysing massive data sets, bioinformatics and biomedicine, artificial intelligence, computational social choice, cognitive science, ecology, and other disciplines.
Research Coordinator: Charles Yeo.