PhD Defense: "High Dimensional Polynomial Approximation with Applications in Imaging and Recognition"

Abhejit Rajagopal

September 5th (Thursday), 11:00am
Harold Frank Hall (HFH), Rm 1132 (CS Conf. Rm.)

Deep learning has demonstrated unreasonable effectiveness on several high dimensional regression and classification problems, far exceeding theoretical expectations. In this talk, we analyze this phenomena from the perspective of approximation theory. Utilizing recent theoretical advances, we investigate whether and under what conditions deep networks can escape the curse of dimensionality, providing experimental evidence where the theory falls short. We use these insights to suggest new approaches to network design that is more in accordance with this theory, and give several examples of where such designs succeed.

About Abhejit Rajagopal:

Abhejit is a PhD candidate in the Scientific Computing Group, advised by Dr. Shivkumar Chandrasekaran. He holds a M.S. in Electrical & Computer Engineering (2016) from UC Santa Barbara, and a B.S. in Electrical Engieneering (2014) from UCLA. His research activities include developing new learning-based approaches to image reconstruction and image recognition in a variety of signal modalities (ECG, EEG, EO/IR, RaDAR, LiDAR, XRCT). His research is funded by multiple grants from the Air Force and Navy, for which he is the principal investigator. Following completion of his degree, he will start as a postdocoral scholar in the Department of Radiology at UC San Francisco, developing machine learning techniques for MRI imaging and assessment.

Hosted by: Professor Shivkumar Chandrasekaran