2025/05/23 by Reda Chhaibi, Chhaibi, Reda, Serge Gratton +3
Chemistry · Physics and Astronomy · #Computation #Computation (stat.CO) #Convergence (economics) #Exponential function #Exponential growth #FOS: Computer and information sciences #FOS: Mathematics #Machine Learning (stat.ML) #Numerical Analysis (math.NA) #Optimization and Control (math.OC) #Perturbation (astronomy) #Probability (math.PR) #Quantum optics and atomic interactions #Regularization (linguistics) #Spectroscopy and Laser Applications
paper · pdf · doi:10.48550/arxiv.2506.14780
openalex publication_date 2025/05/23 · openalex created_date 2025/10/19 · openalex updated_date 2026/08/05
In this paper, we propose an accelerated version for the Sinkhorn algorithm, which is the reference method for computing the solution to Entropic Optimal Transport. Its main draw-back is the exponential slow-down of convergence as the regularization weakens ε → 0. Thanks to spectral insights on the behavior of the Hessian, we propose to mitigate the problem via an original spectral warm-start strategy. This leads to faster convergence compared to the reference method, as also demonstrated in our numerical experiments.