UTS site search

Dr Wei Liu


Dr Wei Liu is a Lecturer and Program Leader in the Data Science research program at the Advanced Analytics Institute, School of Software, Faculty of Engineering and IT, the University of Technology, Sydney (UTS). Before joining UTS, he was a Data Mining Research Fellow at the University of Melbourne, and then an Industry-focused Machine Learning Researcher working in the transportation industry at National ICT Australia (NICTA). He obtained his PhD degree in Data Mining Research from the University of Sydney (USYD).

His research outputs are mostly published in journals and conferences that are ranked at A* and A (i.e., top-prestige) by ARC ERA 2010 ranking and by Core academic ranking. He has received 3 Best Paper Awards.

Dr Liu has led several major research projects in partnership with government agencies and industrial organisations in Internet Security, Insurance, Capital Market, Transportation, and Infrastructure sectors. He has developed cutting-edge data mining methods and software tools for the transport industry, which accurately identify causes of road incidents from dynamic traffic networks. He has also designed advanced predictive analysis models for the problems of rare event prediction, fraud/intrusion detection, and spam filtering with time-evolving data distributions, which are validated and applied in different industrial organisations. Details of some of his research projects are in the below:

•  “Traffic Watch for Transport Control Service”, industry partner: Transport Management Centre; May 2013 – June 2014.

•  “Congestion Propagation and Hotspot Detection in Sydney CBD”, industry partner: NSW RMS; Aug – Dec 2013.

•  “Data Fusion Technologies for Comprehensive Transport Data Analysis in Melbourne”, industry partner: VicRoads; Jun – Sep 2013.

•  “Time of Arrival Estimations using HD Vehicle Trajectories”, industry partner: Tomtom. Jan 2013 – March 2013.

•  “Early Detection of Road Traffic Incidents using Social Media”, industry partner: the Transport Management Centre; Oct – Dec 2012.

•  “Causal Inference for Sequential Traffic Congestion", industry partner: Microsoft Research Asia; Nov 2010 – Mar 2011.

•  “Abnormal Claim Detection from Worker’s Compensations”, industry partner: CGU Insurance; Mar 2010 – Jun 2011.

•  “Data Integration for Cross-Market Capital Trading Systems”, industry partner: the SMARST Group (now purchased by Nasdaq), Jun 2008 – Dec 2009.

Image of Wei Liu
Lecturer, A/DRsch Advanced Analytics Institute
Member, Institute of Electrical and Electronics Engineers
Member, Association for Computing Machinery
+61 2 9514 3782

Research Interests

Main Research Interests:

  • Graph mining, network analysis, tensor factorization
  • Causal inference, Granger causality
  • Game theoretical modeling, adversarial learning 
  • Data imbalance learning, cost-sensitive learning
  • Anomaly (outlier) detection
Can supervise: Yes

Competitive PhD scholarships are available for prospective local and international research students.

Data Mining and Knowledge Discovery; Data Analytics.


Liu, W. & Williams, M. 2008, 'Strategies for Business in Virtual Worlds: Case Studies in Second Life', Pacific Asia Conference on Information Systems, PACIS 2008, City University of Hong Kong Press, China, pp. 888-900.
In this paper, we use qualitative and quantitative methodologies to analyse and understand a range of business strategies in Second Life.
Xu, K., Chen, X., Liu, W. & Williams, M. 2006, 'Legged Robot Gait Locus Generation Based on Genetic Algorithms', International Symposium on Practical Cognitive Agents and Robots (PCAR 2006) - Proceedings, ACM digital library, Australia, pp. 51-62.
View/Download from: Publisher's site
Achieving an effective gait locus for legged robots is a challenging task. It is often done manually in a laborious way due to the lack of research in automatic gait locus planning. Bearing this problem in mind, this article presents a gait locus planning method using inverse kinematics while incorporating genetic algorithms. Using quadruped robots as a platform for evaluation, this method is shown to generate a good gait locus for legged robots.

Journal articles

Liu, W., Sukhorukov, A., Miroshnichenko, A., Poulton, C., Xu, Z., Neshev, D. & Kivshar, Y. 2010, 'Complete spectral gap in coupled dielectric waveguides embedded into metal', Appl. Phys. Lett., vol. 97, p. 021106.
View/Download from: Publisher's site
We study a plasmonic coupler involving backward (TM_01) and forward (HE_11) modes of dielectric waveguides embedded into infinite metal. The simultaneously achievable contradirectional energy flows and codirectional wavevectors in different channels lead to a spectral gap, despite the absence of periodic structures along the waveguide. We demonstrate that a complete spectral gap can be achieved in a symmetric structure composed of four coupled waveguides.
Chen, X., Liu, W. & Williams, M. 2009, 'Introduction: Practical Cognitive Agnets and Robots', Autonomous Agents And Multi-Agent Systems, vol. 19, no. 3, pp. 245-247.
Liu, W. & Williams, M. 2002, 'Trustworthiness Of Information Sources And Information Pedigrees', Lecture Notes In Computer Science vol 2333 - Intelligent Agents Viii: Agent Theories, Architectures, And Languages, vol. 2333, pp. 290-306.
View/Download from: Publisher's site
To survive, and indeed thrive, in an open heterogenous information sharing environment, an agent's ability to evaluate the trustworthiness of other agents becomes crucial. In this paper, we investigate a procedure for evaluating an agent's trustworthines