Dimension‐independent Markov chain Monte Carlo on the sphere

We consider Bayesian analysis on high-dimensional spheres with angular central Gaussian priors. These priors model antipodally symmetric directional data, are easily defined in Hilbert spaces and occur, for instance, density estimation binary level set inversion. In this paper we derive efficient Markov chain Monte Carlo methods approximate sampling of posteriors respect to these Our approaches rely lifting the problem ambient space exploit existing dimension-independent samplers linear spaces. By a push-forward kernel construction then obtain chains sphere which inherit reversibility spectral gap properties from Moreover, our proposed algorithms show efficiency numerical experiments. This article is protected by copyright. All rights reserved.