Variable Selection in Maximum Mean Discrepancy for Interpretable Distribution Comparison

no code implementations • 2 Nov 2023 • Kensuke Mitsuzawa, Motonobu Kanagawa, Stefano Bortoli, Margherita Grossi, Paolo Papotti

For this optimisation, we introduce sparse regularisation and propose two methods for dealing with the issue of selecting an appropriate regularisation parameter.

Causal Inference Two-sample testing +1