Skip to main content

Machine Learning and Data Mining

W. Liu, X. Cao, C. P. de Campos, J. Hong, B. Murphy

Knowledge acquisition is expensive and often there is no expert around from whom to elicit the knowledge. We study Machine Learning and Data Mining techniques that allow models that are easily comprehensible to humans to be constructed from raw data.

Our earliest work in this area started with developing learning algorithms for constructing Bayesian Networks from data. Our current research in this area is strongly influenced by the need to address global societal challenges in the Big Data era.

  1. Developing Graph-based pattern and anomaly detection algorithms with applications to cyber-physical systems security in CSIT; exercise patterns and influences discovery among participants in physical activity schemes in collaborating with the Centre for Public Health; social connection patterns discovery from social networks.
  2. Developing various streaming data analytical approaches to analyzing real-time data from sensor and social networks (e.g. smart grids, transport, Twitter messages). Examples include collaboration with Belfast City Council on analyzing data on Pollution, Waste disposal, Treatment, and Recycling; Anti-Social Behaviours, etc.
  3. Developing real-time threats and anomaly monitoring and prediction algorithms, using knowledge discovered above, to provide real-time situation awareness for decision support.
  4. Developing graphical models for learning knowledge and making inferences from data, especially with incomplete and uncertain datasets, applications include genomic data and image data.
  5. Developing machine learning approaches to analyzing brain signals (e.g. EEG data) for discovering patterns or early signs of diseases.