Kabir Aladin Verchand

I am a postdoctoral scholar hosted jointly by Richard Samworth at the Statistical Laboratory, University of Cambridge and Ashwin Pananjady at the Schools of Industrial and Systems Engineering and Electrical and Computer Engineering, Georgia Tech. Previously, I obtained my PhD in Electrical Engineering at Stanford University, where I was advised by Andrea Montanari.

I am broadly interested in problems at the intersection of optimization, statistics, and probability. In particular, I am interested in understanding both statistical and computational aspects of learning from random (as opposed to worst case) data. Most of my recent work has focused on providing sharp, algorithm-specific guarantees for various nonconvex optimization problems with random data.

Note: I previously published under the name Kabir Aladin Chandrasekher.

Selected and recent publications

1. Chandrasekher, K.A., Lou, M., and Pananjady, A. (2022), Alternating minimization for generalized rank one matrix sensing: Sharp predictions from a random initialization
   • Extended abstract at Algorithmic Learning Theory (ALT) 2024.
2. Chandrasekher, K.A., Pananjady, A., and Thrampoulidis, C. (2023), Sharp global convergence guarantees for iterative nonconvex optimization: A Gaussian process perspective (Annals of Statistics).
   • Runner-up: Best paper prize for young researchers in continuous optimization (ICCOPT 2022)

3. Lou, M., Verchand, K.A., and Pananjady, A. (2024), Hyperparameter tuning via trajectory predictions: Stochastic prox-linear methods in matrix sensing

4. Mardia, J., Verchand, K.A., and Wein, Alexander S. (2024), Low-degree phase transitions for detecting a planted clique in sublinear time