Tong Xin

Tong Xin is Associate Professor in the Department of Mathematics at National University of Singapore. His research develops theoretical frameworks for high-dimensional stochastic systems, focusing on algorithmic stability in optimization and sampling methods.

Core research areas include dimension-independent convergence guarantees for stochastic gradient descent, efficiency of ensemble Kalman methods in inverse problems, and replica exchange techniques for non-convex optimization. Recent work establishes robustness bounds for Markov Chain Monte Carlo variants and analyzes generalization errors in machine learning.

Applications span uncertainty quantification in physical systems, statistical inference for non-i.i.d. data streams, and computational methods for large-scale inverse problems. Editorial roles include SIAM Journal of Uncertainty Quantification and Statistics and Computing.