Independence Test for Linear Non-Gaussian Data and Applications in Causal Discovery
A kernel-based independence test tailored to linear non-Gaussian models, with applications to causal discovery.
My full publications list is on Google Scholar.
* denotes equal contributions and † denotes corresponding authors.
A kernel-based independence test tailored to linear non-Gaussian models, with applications to causal discovery.
A conditional-ICA framework for identifying latent-variable causal structures without relying on purity assumptions.
Adaptive sample reweighting exposes rare dependence patterns that conventional global tests can miss.
A constrained-optimization framework for enforcing interventional fairness when the causal graph is only partially known.