TODO

Research Interests

I am interested in machine learning in biomedical domains. My undergraduate and master's degrees are in Computer Science, where I focused on machine learning methods. Now, I'm working on multi-view learning to predict survival outcomes in cancer.

Overlapping Group LASSO

With Yulia Newton, I'm working on an R package for overlapping group LASSO. We would like to be able to use LASSO with pathway analysis.

Multi-view Learning

My favorite project. Using multiple, redundant, views to help with survival predictions in different cancer types. Currently I'm focusing on imaging and genomic data in TCGA GBM.

DREAM8 Challenge 1A Winners

TODO

TODO: Posters

2014 AACR

We identified pathways relevant for metastatic site prediction in prostate cancer, using multi-class linear predictors.

RECOMB: DREAM8 1A Method

A poster of our method for the DREAM challenge, presented at the PBSE recruitment event.

RMCWIC: Best Poster Award

I presented my rotation project in Ed Green's lab, on converting MIA to SAM format and the importance of aligners specifically for ancient DNA.