Research

In general, I am interested in the connection between machine learning and human learning. More specifically, I am interested in how machine learning methods can inform us in the effort to understand human learning, and how phenomena in human learning can inspire machine learning algorithms. To that end, I am currently investigating novel data analysis methods for fMRI data that can potentially reveal the mechanisms of the brain. I have worked on a method to automatically infer cognitive processes from fMRI data, and now I am working on methods to combine data from multiple domains (e.g. subjects and studies). The group's Web site is here.

My advisor is Tom Mitchell.

Thesis

Publications

Integrating Multiple-Study Multiple-Subject fMRI Datasets Using Canonical Correlation Analysis (with Marcel Adam Just and Tom M. Mitchell), Proceedings of the MICCAI 2009 Workshop: Statistical modeling and detection issues in intra- and inter-subject functional MRI data analysis, 2009.

Modeling fMRI data generated by overlapping cognitive processes with unknown onsets using Hidden Process Model (with Rebecca A. Hutchinson, Radu Stefan Niculescu, Timothy A. Keller, and Tom M. Mitchell), NeuroImage 46(1), 2009.

Classifying Multiple-Subject fMRI Data Using the Hierarchical Gaussian Naïve Bayes Classifier, 13th Conference on Human Brain Mapping, 2007.

Hierarchical Gaussian Naïve Bayes Classifier for Multiple-Subject fMRI Data, New Directions on Decoding Mental States from fMRI Data, NIPS Workshop, 2006.

Hidden Process Models (with Rebecca Hutchinson and Tom Mitchell), 23rd International Conference on Machine Learning, 2006.

Learning to Identify Overlapping and Hidden Cognitive Processes from fMRI Data (with Rebecca Hutchinson and Tom Mitchell), 11th Conference on Human Brain Mapping, 2005.

Presentations

Feature Selection on Raw and Wavelet-Transformed fMRI Data, CNBC Retreat, October 2005.