新的传输滤波方法改进了非线性、非高斯近似

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-13 08:36

研究人员开发了一种新的无似然传输滤波方法，该方法利用状态变量和观测变量之间的耦合。该方法将滤波分析步骤重新表述为最小化真实联合测度和近似联合测度之间的最大均值差异（MMD）。该方法提供了传输图的解析计算，避免了粒子坍塌，并能准确近似非高斯滤波后验，在非线性、非高斯场景中表现出优越的性能。 AI

影响引入了一种新颖的统计方法来近似复杂的概率分布，有可能改进那些依赖于动态环境中准确状态估计的AI系统。

排序理由该集群包含一篇详细介绍新统计方法的学术论文。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv stat.ML TIER_1 English(EN) · Dengfei Zeng, Lijian Jiang, Shuyu Sun, Dunhui Xiao · 2026-05-14 04:00

Coupling-Informed Transport Maps for Bayesian Filtering in Nonlinear Dynamical Systems

arXiv:2605.13174v1 Announce Type: new Abstract: A likelihood-free transport filtering method is proposed based on the couplings between state and observation variables. By exploiting a block-triangular structure in the transport map, the analysis step of filtering is reformulated…
arXiv stat.ML TIER_1 English(EN) · Dunhui Xiao · 2026-05-13 08:36

Coupling-Informed Transport Maps for Bayesian Filtering in Nonlinear Dynamical Systems

A likelihood-free transport filtering method is proposed based on the couplings between state and observation variables. By exploiting a block-triangular structure in the transport map, the analysis step of filtering is reformulated as the minimization of the maximum mean discrep…