Markovian Fragmentation

Geometry-driven Bayesian Inference for Ultrametric Covariance Matrices
We conduct the Bayesian inference that makes efficient local move on the space of ultrametric matrices by leveraging the bijection map between the ultrametric matrices and the tree space.
Tsung-Hung Yao, Zhenke Wu, Karthik Bharath, Veerabhadran Baladandayuthapan
PDF Code
Geometry-driven Bayesian Inference for Ultrametric Covariance Matrices