Semi-Linear Gaussian Bayesian Networks via Penalized B-Splines
October 23, 2026
SLGBN models bridge the gap between restrictive linear Gaussian networks and computationally expensive non-parametric models. The framework uses penalized B-splines within generalized additive models to capture nonlinear conditional mean functions while preserving Gaussian distributions.
HOW THIS AFFECTS YOU
●
researcherYou can model nonlinear dependencies in continuous variables more efficiently using this hybrid approach.