{ "id": "2410.17297", "version": "v1", "published": "2024-10-22T07:38:23.000Z", "updated": "2024-10-22T07:38:23.000Z", "title": "Error estimates between SGD with momentum and underdamped Langevin diffusion", "authors": [ "Arnaud Guillin", "Yu Wang", "Lihu Xu", "Haoran Yang" ], "categories": [ "stat.ML", "cs.LG", "math.PR" ], "abstract": "Stochastic gradient descent with momentum is a popular variant of stochastic gradient descent, which has recently been reported to have a close relationship with the underdamped Langevin diffusion. In this paper, we establish a quantitative error estimate between them in the 1-Wasserstein and total variation distances.", "revisions": [ { "version": "v1", "updated": "2024-10-22T07:38:23.000Z" } ], "analyses": { "keywords": [ "underdamped langevin diffusion", "stochastic gradient descent", "total variation distances", "close relationship", "popular variant" ], "note": { "typesetting": "TeX", "pages": 0, "language": "en", "license": "arXiv", "status": "editable" } } }