利用分数匹配目标训练基于能量的归一化流
Training Energy-Based Normalizing Flow with Score-Matching Objectives
AI总结：
本文提出基于能量的归一化流（EBFlow），利用分数匹配目标训练以绕过雅可比行列式计算，从而支持任意线性层并提升训练效率，实验表明其在速度和负对数似然上均优于最大似然方法。
摘要：
本文建立了基于流和基于能量的生成模型参数化之间的联系，并提出了一种新的基于流的建模方法，称为基于能量的归一化流（EBFlow）。我们证明，通过使用分数匹配目标优化EBFlow，可以完全绕过线性变换的雅可比行列式计算。这一特性使得在构建基于流的模型时可以使用任意线性层，而不会将每个训练迭代的计算时间复杂度从$O(D^2L)$增加到$O(D^3L)$，其中模型具有$L$层并接受$D$维输入。这使得EBFlow的训练比常用的最大似然训练方法更高效。除了运行时间的减少，我们还基于对分数匹配方法的分析开发了多种技术，以增强EBFlow的训练稳定性和经验性能。实验结果表明，与最大似然估计相比，我们的方法实现了显著的加速，同时在负对数似然（NLL）方面以明显优势优于先前的方法。
英文摘要：
In this paper, we establish a connection between the parameterization of flow-based and energy-based generative models, and present a new flow-based modeling approach called energy-based normalizing flow (EBFlow). We demonstrate that by optimizing EBFlow with score-matching objectives, the computation of Jacobian determinants for linear transformations can be entirely bypassed. This feature enables the use of arbitrary linear layers in the construction of flow-based models without increasing the computational time complexity of each training iteration from $O(D^2L)$ to $O(D^3L)$ for an $L$-layered model that accepts $D$-dimensional inputs. This makes the training of EBFlow more efficient than the commonly-adopted maximum likelihood training method. In addition to the reduction in runtime, we enhance the training stability and empirical performance of EBFlow through a number of techniques developed based on our analysis of the score-matching methods. The experimental results demonstrate that our approach achieves a significant speedup compared to maximum likelihood estimation while outperforming prior methods with a noticeable margin in terms of negative log-likelihood (NLL).
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
