Learning to Draw Samples: With Application to Amortized MLE for Generative Adversarial Learning
This addresses the challenge of efficient sampling for probabilistic inference, with applications in generative modeling, though it builds incrementally on existing variational and adversarial methods.
The paper tackles the problem of training stochastic neural networks to sample from target distributions by adjusting parameters along Stein variational gradients to minimize KL divergence, and applies this to amortized MLE for deep energy models, achieving competitive image generation results.
We propose a simple algorithm to train stochastic neural networks to draw samples from given target distributions for probabilistic inference. Our method is based on iteratively adjusting the neural network parameters so that the output changes along a Stein variational gradient that maximumly decreases the KL divergence with the target distribution. Our method works for any target distribution specified by their unnormalized density function, and can train any black-box architectures that are differentiable in terms of the parameters we want to adapt. As an application of our method, we propose an amortized MLE algorithm for training deep energy model, where a neural sampler is adaptively trained to approximate the likelihood function. Our method mimics an adversarial game between the deep energy model and the neural sampler, and obtains realistic-looking images competitive with the state-of-the-art results.