Reweighted wake-sleep (RWS) is a machine learning method for performing Bayesian inference in a very general class of models. RWS draws $K$ samples from an underlying approximate posterior, then uses importance weighting to provide a better estimate of the true posterior. RWS then updates its approximate posterior towards the importance-weighted estimate of the true posterior. However, recent work [Chattergee and Diaconis, 2018] indicates that the number of samples required for effective importance weighting is exponential in the number of latent variables. Attaining such a large number of importance samples is intractable in all but the smallest models. Here, we develop massively parallel RWS, which circumvents this issue by drawing $K$ samples of all $n$ latent variables, and individually reasoning about all $K^n$ possible combinations of samples. While reasoning about $K^n$ combinations might seem intractable, the required computations can be performed in polynomial time by exploiting conditional independencies in the generative model. We show considerable improvements over standard "global" RWS, which draws $K$ samples from the full joint.

翻译：重加权唤醒-睡眠（RWS）是一种在非常广泛的模型类别中执行贝叶斯推理的机器学习方法。RWS从底层近似后验中抽取$K$个样本，然后使用重要性加权来更准确地估计真实后验。接着，RWS将其近似后验向重要性加权估计的真实后验方向更新。然而，近期研究[Chatterjee和Diaconis, 2018]表明，有效重要性加权所需的样本数量与隐变量数量呈指数关系。除最小模型外，获取如此大量的重要性样本在实践中是不可行的。本文中，我们发展了大规模并行RWS，通过抽取全部$n$个隐变量的$K$个样本，并分别推理所有$K^n$种可能的样本组合，来规避这一问题。尽管处理$K^n$种组合看似不可行，但通过利用生成模型中的条件独立性，所需的计算可以在多项式时间内完成。我们展示了与标准的“全局”RWS（从完整联合分布中抽取$K$个样本）相比的显著改进。