Welcome to the Pattern Discovery and Machine Learning Laboratory at UT Dallas!
Our lab is focused on development of novel machine learning methods for scalable and accurate detection of anomalous or interesting patterns (e.g., events, frauds, sub-network markers, storylines, knowledge patterns) in complex, massive, and high-dimensional datasets. By creating, deploying, and evaluating new methods in collaboration with public sector partners, we hope both to advance the state of the art in machine learning and to improve the quality of public health, safety, and security.
The lab is currently directed by Prof. Feng Chen. Please feel free to reach out!
Research at PDML
The lab's three main research areas include:
- Pattern detection and prediction. We are interested in developing meta-learning and few-shot learning of machine learning/deep learning models for detecting and forecasting patterns with a minimal set of training examples. We are particularly interested in meta-learning of deep learning models in a high-dimensional, dynamic, and complex environment, where the data distribution may evolve over time.
- Pattern uncertainty and interpretation. We are interested in examining multiple dimensions of pattern uncertainty and interpretation derived from heterogeneous root causes and investigating the synergistic merits of uncertainty research explored in both belief/evidence theory (e.g., Dempster-Shafer Theory, Subjective Logic) and ML/DL.
- Pattern fairness and causality. We are interested in developing fair machine learning systems and understanding the causes of unfairness for various pattern discovery tasks. We are particularly interested in meta-learning and causal learning of online fair deep learning models a high-dimensional, dynamic, and complex environment.
Key application areas include:
- Public health and disease surveillance
- Cyber attack detection and prevention
- Crime prediction and prevention
- Road congestion and incident detection and forecasting
- Sociental event forecasting in social media
- Fairness in criminal justice
- Fraud detection in finance
- Environmental health and prevention
Additional details on many of these projects will be posted soon; in the meantime, please feel free to browse some of our recent publications.