arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.20482cs.AI
收藏

R-ParVI:基于奖励视角的粒子变分推断

R-ParVI: Particle-based variational inference through lens of rewards

Yongchao Huang

AI总结:

R-ParVI 是一种奖励驱动、无梯度的粒子变分推断方法,通过粒子流采样部分已知密度,兼顾高密度区域探索与多样性保持,适用于贝叶斯推断和生成建模。

摘要:

本文提出了一种奖励引导、无梯度的粒子变分推断(ParVI)方法,称为 R-ParVI,用于对部分已知密度(例如,相差一个常数)进行采样。R-ParVI 将采样问题表述为由奖励驱动的粒子流:粒子从先验分布中抽取,在参数空间中移动,其移动由融合了目标密度评估的奖励机制决定,稳态粒子配置近似目标几何形状。粒子与环境的交互通过随机扰动和奖励机制模拟,驱动粒子向高密度区域移动,同时保持多样性(例如,防止坍缩成簇)。R-ParVI 为一类概率模型(如贝叶斯推断和生成建模中遇到的模型)提供了快速、灵活、可扩展和随机的采样与推断。

英文摘要:

A reward-guided, gradient-free ParVI method, \textit{R-ParVI}, is proposed for sampling partially known densities (e.g. up to a constant). R-ParVI formulates the sampling problem as particle flow driven by rewards: particles are drawn from a prior distribution, navigate through parameter space with movements determined by a reward mechanism blending assessments from the target density, with the steady state particle configuration approximating the target geometry. Particle-environment interactions are simulated by stochastic perturbations and the reward mechanism, which drive particles towards high density regions while maintaining diversity (e.g. preventing from collapsing into clusters). R-ParVI offers fast, flexible, scalable and stochastic sampling and inference for a class of probabilistic models such as those encountered in Bayesian inference and generative modelling.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑