标题： GANs 隐式执行近似贝叶斯模型选择 显示英文标题

标题： GANs Secretly Perform Approximate Bayesian Model Selection

摘要： 生成对抗网络（GANs）是流行且成功的生成模型。 尽管取得了成功，优化过程却以极具挑战性著称，并且它们需要正则化以防止过拟合。 在本工作中，我们通过将GANs解释为概率生成模型来解释其成功之处和局限性。 这种解释使我们能够将GANs视为具有部分随机性的贝叶斯神经网络，从而建立通用近似的条件。 然后，我们可以将几种GAN变体的对抗式优化视为对边缘似然的代理进行优化。 利用边缘似然优化与奥卡姆剃刀之间的联系，我们可以定义正则化和优化策略，以平滑损失景观并寻找具有最小描述长度的解决方案，这些解决方案与平坦的极小值和良好的泛化能力相关。 在广泛实验中的结果表明，这些策略带来了性能提升，并为更深入理解GANs的正则化策略铺平了道路。 显示英文摘要

摘要： Generative Adversarial Networks (GANs) are popular and successful generative models. Despite their success, optimization is notoriously challenging and they require regularization against overfitting. In this work, we explain the success and limitations of GANs by interpreting them as probabilistic generative models. This interpretation enables us to view GANs as Bayesian neural networks with partial stochasticity, allowing us to establish conditions of universal approximation. We can then cast the adversarial-style optimization of several variants of GANs as the optimization of a proxy for the marginal likelihood. Taking advantage of the connection between marginal likelihood optimization and Occam's razor, we can define regularization and optimization strategies to smooth the loss landscape and search for solutions with minimum description length, which are associated with flat minima and good generalization. The results on a wide range of experiments indicate that these strategies lead to performance improvements and pave the way to a deeper understanding of regularization strategies for GANs.