Graduated with honors from Columbia University in 2007 majoring in Mathematics and Economics
I currently research multilabel classification and large scale neural networks for natural language processing, with application to medical data. I also study evaluation methodology with an eye towards philosophical considerations of machine learning. What are useful performance metrics / loss functions. Further, what are we training our systems to do and how does it accord with our intuition about how we think they should ideally behave?
Optimal Thresholding of Classiﬁers to Maximize F1 Measure - Forthcoming at the 2014 European Conference on Machine Learning.