Measure Transport with Kernel Stein Discrepancy
arXiv:2010.11779
Abstract
Measure transport underpins several recent algorithms for posterior approximation in the Bayesian context, wherein a transport map is sought to minimise the Kullback--Leibler divergence (KLD) from the posterior to the approximation. The KLD is a strong mode of convergence, requiring absolute continuity of measures and placing restrictions on which transport maps can be permitted. Here we propose to minimise a kernel Stein discrepancy (KSD) instead, requiring only that the set of transport maps is dense in an sense and demonstrating how this condition can be validated. The consistency of the associated posterior approximation is established and empirical results suggest that KSD is competitive and more flexible alternative to KLD for measure transport.
References in corpus (23)
- Auto-Encoding Variational Bayes
- An overview of gradient descent optimization algorithms
- NICE: Non-linear Independent Components Estimation
- Density estimation using Real NVP
- A Conceptual Introduction to Hamiltonian Monte Carlo
- Markov Chain Monte Carlo and Variational Inference: Bridging the Gap
- Normalizing Flows for Probabilistic Modeling and Inference
- MADE: Masked Autoencoder for Distribution Estimation
- Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm
- Training generative neural networks via Maximum Mean Discrepancy optimization
- Measuring Sample Quality with Stein's Method
- A Kernel Test of Goodness of Fit
- Large sample analysis of the median heuristic
- Operator Variational Inference
- Optimizing The Integrator Step Size for Hamiltonian Monte Carlo
- Flows for simultaneous manifold learning and density estimation
- Stein Neural Sampler
- Quasi-Monte Carlo Variational Inference
- Invertible Generative Modeling using Linear Rational Splines
- MetFlow: A New Efficient Method for Bridging the Gap between Markov Chain Monte Carlo and Variational Inference
- Manifold lifting: scaling MCMC to the vanishing noise regime
- Blindness of score-based methods to isolated components and mixing proportions
- Variational Mixture of Normalizing Flows