Rajiv Khanna
Email: [first name]ak@berkeley.edu
Google Scholar

I am currently a Visiting Faculty Researcher at Google. In Spring 2022, I will be joining Department of Computer Science at Purdue University as an Assistant Professor!

Previously, I was a Postdoctoral Scholar at the Foundations of Data Analystics Institute at University of California, Berkeley and am very fortunate to be working with Michael Mahoney. Before that, I was a Research Fellow in the Foundations of Data Science program at the Simons Institute in Fall 2018. I graduated with my PhD from UT Austin. I am very fortunate to be able to learn from and interact with Professors Joydeep Ghosh, and Alex Dimakis. My PhD research is to study various aspects of greedy algorithms in machine learning, both from theory and practical viewpoints. From the theory side, I have worked on studying approximation guarantees for "greedy-like" algorithms for sparsity and rank constrained problems for general functions based on their smoothness and convexity, and convergence rates for greedy algorithms like accelerated IHT, Matching Pursuit and boosting . From practical side, I have worked on using greedy selections for approximate variational inference for sparse regression, sparse PCA for fMRI applications, and for interpreting black box models. In the past I have worked on problems involving predicting buying propensity based on marketing touches, large scale recommendation systems, Ad Click prediction, and ranking.