Welcome. I am a PhD student at the University of Texas at Austin. Currently, my research includes work on anomaly detection and classifying imbalanced datasets. I am also interested in text mining. Below are some of my publications.

---

Resume (in PDF format)
Alex Liu's resume

---

Publications

• AI Lessons Learned from Experiments in Insider Threat Detection
  • Accepted to AAAI 2006 Spring Symposium (What Went Wrong and Why: Lessons from AI Research and Applications)
  • Joint work with Applied Research Laboratories, University of Texas
  • Description: We discuss some lessons learned about applying machine learning techniques to new domains by using experiments in insider threat detection as a running example.

• A Comparison of System Call Feature Representations for Insider Threat Detection
  • Accepted to IEEE 2005 Information Assurance Workshop
  • Joint work with Applied Research Laboratories, University of Texas
  • Description: In this paper, we attempt to detect when authorized users abuse their priviledges via supervised outlier detection.

• Master's Thesis
  • Description: My thesis work addresses the problem of classifying highly imbalanced datasets (e.g. when the prior probability of one class is much lower than the prior probability of the other class). In particular, I focus on text datasets since most previous papers focus on low dimensional data. I introduce a number of new techniques to handle class imbalance and benchmark these new techniques and previously published techniques on seven text datasets.