Sampling from multi-modal distributions with polynomial query complexity in fixed dimension via reverse diffusion

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Vacher, Adrien, Chehab, Omar, Korba, Anna
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914109177462784
author Vacher, Adrien
Chehab, Omar
Korba, Anna
author_facet Vacher, Adrien
Chehab, Omar
Korba, Anna
contents Even in low dimensions, sampling from multi-modal distributions is challenging. We provide the first sampling algorithm for a broad class of distributions -- including all Gaussian mixtures -- with a query complexity that is polynomial in the parameters governing multi-modality, assuming fixed dimension. Our sampling algorithm simulates a time-reversed diffusion process, using a self-normalized Monte Carlo estimator of the intermediate score functions. Unlike previous works, it avoids metastability, requires no prior knowledge of the mode locations, and relaxes the well-known log-smoothness assumption which excluded general Gaussian mixtures so far.
format Preprint
id arxiv_https___arxiv_org_abs_2501_00565
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Sampling from multi-modal distributions with polynomial query complexity in fixed dimension via reverse diffusion
Vacher, Adrien
Chehab, Omar
Korba, Anna
Computation
Machine Learning
Statistics Theory
Even in low dimensions, sampling from multi-modal distributions is challenging. We provide the first sampling algorithm for a broad class of distributions -- including all Gaussian mixtures -- with a query complexity that is polynomial in the parameters governing multi-modality, assuming fixed dimension. Our sampling algorithm simulates a time-reversed diffusion process, using a self-normalized Monte Carlo estimator of the intermediate score functions. Unlike previous works, it avoids metastability, requires no prior knowledge of the mode locations, and relaxes the well-known log-smoothness assumption which excluded general Gaussian mixtures so far.
title Sampling from multi-modal distributions with polynomial query complexity in fixed dimension via reverse diffusion
topic Computation
Machine Learning
Statistics Theory
url https://arxiv.org/abs/2501.00565