Partially Linear Additive Gaussian Graphical Models
[edit]
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:21802190, 2019.
Abstract
We propose a partially linear additive Gaussian graphical model (PLAGGM) for the estimation of associations between random variables distorted by observed confounders. Model parameters are estimated using an $L_1$regularized maximal pseudoprofile likelihood estimator (MaPPLE) for which we prove a $\sqrt{n}$sparsistency. Importantly, our approach avoids parametric constraints on the effects of confounders on the estimated graphical model structure. Empirically, the PLAGGM is applied to both synthetic and realworld datasets, demonstrating superior performance compared to competing methods.
Related Material


