3 papers
math.NA2025
A federated Kaczmarz algorithm
Halyun Jeong, Deanna Needell, Chi-Hao Wu
In this paper, we propose a federated algorithm for solving large linear systems that is inspired by the classic randomized Kaczmarz algorithm. We provide convergence guarantees of…
cs.LG2024
Robust Fourier Neural Networks
Halyun Jeong, Jihun Han
Fourier embedding has shown great promise in removing spectral bias during neural network training. However, it can still suffer from high generalization errors, especially when th…
cs.LG2024
Stochastic gradient descent for streaming linear and rectified linear systems with adversarial corruptions
Halyun Jeong, Deanna Needell, Elizaveta Rebrova
We propose SGD-exp, a stochastic gradient descent approach for linear and ReLU regressions under Massart noise (adversarial semi-random corruption model) for the fully streaming se…