Peer-reviewed papers and preprints in optimization, bandits, and graph theory. Where the code is public, it's linked directly.
P Thaker (Mitsubishi Electric Research Laboratories)
A control system that directs teams of search agents along fuel-constrained paths using a multi-level multi-armed bandit, classifying regions from noisy measurements with probabilistic finite-time guarantees.
P. Thaker, V. Gattani, V. Tirukkonda, P. Saidi et al.
A multi-armed bandit method that uses the Ramanujan periodicity transform to detect periodic patterns in shifting rewards and exploit them to reduce regret.
P. Thaker, S. Di Cairano, A. P. Vinod
A thresholding multi-armed bandit algorithm for coordinating multiple search agents under noisy sensor data, with finite bounds on search time, labeling time, and cost.
P Thaker, M Malu, N Rao, G Dasarathy
A pure-exploration bandit method (GRUB and ζ-GRUB) that uses a similarity graph over arms to find the best arm, or one above a threshold, with far fewer samples than standard UCB.
J Janiczek, P Thaker, G Dasarathy, C S Edwards et al.
Makes a physics-based optical dispersion model differentiable and uses it for end-to-end hyperspectral unmixing, jointly estimating material abundances and physical parameters.
P. K. Thaker, G. Dasarathy, A. Nedic
Proves identifiability and sample-complexity bounds for recovering a complex vector from quadratic measurements, and shows the nonconvex landscape lets gradient descent converge globally from arbitrary initialization.
P. Thaker, A. Gopalan, R. Vaze
A distributed learning algorithm that lets players competing for an intermittently-available server choose when to sample it, converging to a unique Nash equilibrium while keeping sensing costs low.