Latest Transcript: pdf
Technologies: Python, Git, Javascript, Bash, Cython Data Science Libraries: NumPy, pandas, scikit-learn, Keras, TensorFlow, Cython
- Example of E2E machine learning project (Fiddler Labs system used to Attribute Drifts in Production Data Streams) - machine learning model training + validation, model serving backend, and hosting website - code, live demo website (Chrome recommended, Safari does not work)
- Generic Software Development Experience - python code, javascript code, bash scripts
- Optimising a Statistical Relational Learning library using Cython - code, final report
- Data Science Project Work - final report
Topics: Influence Maximization, Social Networks; Bradley-Terry Models, Recourse; Counterfactuals, Machine Learning
- Maximizing Conrasting Opinions in Signed Social Networks: paper, code and presentation
- Recourse for the Real World: Learning Feature Modification Costs from Pairwise Comparisons: paper, code
- Beyond Individualized Recourse: Interpretable and Interactive Summaries of Actionable Recourses: paper, code (unavailable due to NDA)
- Can I Still Trust You? Understanding the Impact of Distribution Shifts on Algorithmic Recourses: paper, code
Open source and Community: talks and teaching