Non-parametric Sampling Approximation via Voronoi Tessellations

In this paper we propose a novel non-parametric sampling approach to estimate posterior distributions from parameters of interest. This technique is particularly suited for models that are computationally expensive to evaluate. Starting from an initial sample over the parameter space, this method makes use of this initial information to form a geometrical structure known as Voronoi tessellation over the whole parameter space. This rough approximation to the posterior distribution provides a way to generate new points from the posterior distribution without any additional costly model evaluations. By using a traditional MCMC over the non-parametric tessellation, the initial approximate distribution is refined sequentially, allowing to sample new points at any moment. We applied this method to a couple of climate models to show that this hybrid scheme successfully approximates the posterior distribution of the model parameters without any additional forward evaluation of the model itse...

• Type: Journal
• Year: 2016
• Where: CSSC
• Authors: Alejandro Villagran, Gabriel Huerta, Marina Vannucci, Charles S. Jackson, Alvaro Nosedal
• Keywords: Automated Reasoning, CSSC 2016