Trustworthy Machine Learning
Building reliable, interpretable, and statistically grounded machine learning methods for high-stakes decision making.
Research
From statistical foundations to deployable machine learning.
Our projects emphasize end-to-end machine learning: data acquisition, feature engineering, modeling, validation, interpretation, and research communication.
Current Research Areas
Our lab studies how machine learning can be more reliable, more useful in real-world settings, and more impactful across scientific domains.
Generative AI & Synthetic Data
Developing principled methods for synthetic data generation, data augmentation, and privacy-preserving machine learning.
Statistical Foundations of AI
Combining statistics and machine learning to improve uncertainty quantification, robustness, model evaluation, and scientific reliability.
The Assembly Line
Our research workflow
Students learn a structured pipeline that moves from raw data to polished research outputs: live demos, posters, technical reports, and manuscripts.
01ExtractionCollect and understand data
02RefineryCleaning, preprocessing, feature engineering
03ForgeMachine learning & statistical modeling
04Quality ControlValidation, robustness, explainability
05DeploymentPapers, posters, software, presentations
Current Research Projects
Reliable Synthetic Data
Trustworthy AI
Adaptive Data Augmentation
AI Uncertainty Quantification
Robust Time Series Forecasting
Machine Learning for Environmental Data
Explainable AI
Statistical Learning Theory
Our students graduate with
✓ Research experience
✓ Conference presentations
✓ Published software
✓ Technical writing
✓ GitHub portfolio
✓ Graduate school preparation
✓ Industry-ready machine learning skills
Technologies
Python
R
TensorFlow
PyTorch
scikit-learn
SQL
Git
Docker
Jupyter
LaTeX