My research focuses on developing and theoretically analyzing new statistical procedures, to learn from the large volume of data available, that are statistically significant, computationally efficient and scientifically meaningful. A particular emphasis is placed on methods with models whose complexity grows with certain characteristics of the dataset (called as non-parametric models in the literature).
Broadly speaking, my contributions add to the set of statistical procedures for extracting meaningful information from large data sets. Part of my work enables one to apply statistical procedures in certain previously unstudied scenarios. Another part of my work, provides novel and useful insights on existing methods that leads to better understanding of such procedures.
I guess the most important aspect of the fellowship is the freedom it provided. It allowed me to visit professors at other universities, share my research and start useful collaborations. It also gave me the opportunity to travel to conferences and present my work. Also, during the visit to Facebook, I had the opportunity to talk to several engineers and get to know the problems they are working on, which gave me a different perspective of the field. It was indeed great being a Facebook fellow.
I am wrapping up my dissertation currently and will soon be starting my postdoc in the statistics department at University of Wisconsin, Madison. I would like to take a high-risk high-yield approach, focusing on specific research problems in the future.