Research Interests:
Broadly speaking, my interests lie in using (or devising new) maching learning methods that result in interpretable models, which can be used to extract biological knowledge from very large data sets. While at Concordia, I worked on a model for the relationship between image intensity and protein copy number in drosophila (see here). Currently I'm working collectively with a group from the Boone and Andrews labs at CCBR on a study of how different genetic deletions from the yeast genome affects yeast cell morphology.
Other interests:
- Semi-supervised learning
- Manifold learning
- Nonparametric Bayesian models
- Spectral clustering
- Kernel methods