Black-box density function estimation using recursive partitioning

10/26/2020
by   Erik Bodin, et al.
0

We present a novel approach to Bayesian inference and general Bayesian computation that is defined through a recursive partitioning of the sample space. It does not rely on gradients, nor require any problem-specific tuning, and is asymptotically exact for any density function with a bounded domain. The output is an approximation to the whole density function including the normalization constant, via partitions organized in efficient data structures. This allows for evidence estimation, as well as approximate posteriors that allow for fast sampling and fast evaluations of the density. It shows competitive performance to recent state-of-the-art methods on synthetic and real-world problem examples including parameter inference for gravitational-wave physics.

READ FULL TEXT
research
11/20/2018

Black-Box Autoregressive Density Estimation for State-Space Models

State-space models (SSMs) provide a flexible framework for modelling tim...
research
05/18/2023

Evidence Networks: simple losses for fast, amortized, neural Bayesian model comparison

Evidence Networks can enable Bayesian model comparison when state-of-the...
research
03/25/2022

LAMBDA: Covering the Solution Set of Black-Box Inequality by Search Space Quantization

Black-box functions are broadly used to model complex problems that prov...
research
08/19/2019

Cumulants of multiinformation density in the case of a multivariate normal distribution

We consider a generalization of information density to a partitioning in...
research
09/14/2020

Density Estimation via Bayesian Inference Engines

We explain how effective automatic probability density function estimate...
research
11/25/2021

Group equivariant neural posterior estimation

Simulation-based inference with conditional neural density estimators is...
research
12/11/2019

CARP: Compression through Adaptive Recursive Partitioning for Multi-dimensional Images

Fast and effective image compression for multi-dimensional images has be...

Please sign up or login with your details

Forgot password? Click here to reset