Stein Neural Sampler
arXiv:1810.03545
Abstract
We propose two novel samplers to generate high-quality samples from a given (un-normalized) probability density. Motivated by the success of generative adversarial networks, we construct our samplers using deep neural networks that transform a reference distribution to the target distribution. Training schemes are developed to minimize two variations of the Stein discrepancy, which is designed to work with un-normalized densities. Once trained, our samplers are able to generate samples instantaneously. We show that the proposed methods are theoretically sound and experience fewer convergence issues compared with traditional sampling approaches according to our empirical studies.
References in corpus (15)
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Large Scale GAN Training for High Fidelity Natural Image Synthesis
- NICE: Non-linear Independent Components Estimation
- Density estimation using Real NVP
- Unrolled Generative Adversarial Networks
- Generative Moment Matching Networks
- How to Escape Saddle Points Efficiently
- Measuring Sample Quality with Stein's Method
- Yes, but Did It Work?: Evaluating Variational Inference
- Measuring Sample Quality with Kernels
- Learning to Draw Samples: With Application to Amortized MLE for Generative Adversarial Learning
- Deep neural networks are robust to weight binarization and other non-linear distortions
- A-NICE-MC: Adversarial Training for MCMC
- On the Expressive Power of Deep Neural Networks
- Operator Variational Inference
Cited by in corpus (11)
- Learning the Stein Discrepancy for Training and Evaluating Energy-Based Models without Sampling
- Sliced Kernelized Stein Discrepancy
- Auxiliary Task Reweighting for Minimum-data Learning
- Posterior Meta-Replay for Continual Learning
- Measure Transport with Kernel Stein Discrepancy
- Active Slices for Sliced Stein Discrepancy
- Variational (Gradient) Estimate of the Score Function in Energy-based Latent Variable Models
- Generative Particle Variational Inference via Estimation of Functional Gradients
- Neural Variational Gradient Descent
- A stepped sampling method for video detection using LSTM
- Adversarial Stein Training for Graph Energy Models