Deep Directed Generative Models with Energy-Based Probability Estimation
Taesup Kim, Yoshua Bengio Department of Computer Science and Operations Research Université de Montréal Montréal, QC, Canada CIFAR Senior Member
Abstract
Training energy-based probabilistic models is confronted with apparently intractable sums, whose Monte Carlo estimation requires sampling from the estimated probability distribution in the inner loop of training. This can be approximately achieved by Markov chain Monte Carlo methods, but may still face a formidable obstacle that is the difficulty of mixing between modes with sharp concentrations of probability. Whereas an MCMC process is usually derived from a given energy function based on mathematical considerations and requires an arbitrarily long time to obtain good and varied samples, we propose to train a deep directed generative model (not a Markov chain) so that its sampling distribution approximately matches the energy function that is being trained. Inspired by generative adversarial networks, the proposed framework involves training of two models that represent dual views of the estimated probability distribution: the energy function (mapping an input configuration to a scalar energy value) and the generator (mapping a noise vector to a generated configuration), both represented by deep neural networks.
中文速览
训练基于能量函数（energy-based model）的概率模型时，最大的拦路虎是必须从模型自身分布中采样来估计梯度，而当概率分布有多个尖锐的峰值时，传统马尔可夫链蒙特卡洛（MCMC）方法极难在不同峰值之间跳转、混合效率极低。为此，这篇论文提出同时训练两个深度神经网络：一个能量函数网络负责给每个样本打分（真实数据能量低、生成数据能量高），一个生成器网络负责把随机噪声直接变换成样本，二者互相"对抗"——能量函数用生成器的样本替代MCMC来估计负相梯度，生成器则被能量函数的梯度引导去产生低能量（更真实）的样本，同时引入一个熵正则化项防止生成器模式崩塌。实验表明，这套框架能在图像等复杂数据上学到高质量的能量模型，同时生成更多样、更真实的样本。这项工作的意义在于，它用一次前向传播就能完成采样，彻底绕开了MCMC的混合难题，为能量模型的大规模训练提供了一条切实可行的新路。
原文 arXiv:1606.03439;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1606.03439v1