Id: arxiv/2608.04430 Url: https://arxiv.org/abs/2608.04430 Tags: math.NA, stat.ML Score: 0 Title: An adaptive split-combine Gaussian mixture filter for nonlinear and multimodal state estimation Domain: math.NA Source: arxiv Authors: San Kim, Won Chang, Daniel B. Forger, Dae Wook Kim Reviews: 0 Subject: math.NA Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-08-05 Page: https://delphi.science/r/arxiv/2608.04430 Filtering combines model predictions with measurements to estimate the probability density function (PDF) of a system state over time. The PDF often becomes highly asymmetric and even multimodal in nonlinear systems with oscillatory or chaotic dynamics. Such non-Gaussian features violate the single-Gaussian assumption underlying Kalman-type filters. To address this problem, Gaussian mixture filtering has been proposed. However, accurately propagating mixture components and adaptively adjusting their number and weights over time remain open challenges. Here, we develop an adaptive…