神经常微分方程
Neural Ordinary Differential Equations
摘要：
我们介绍了一个新的家庭的深度神经网络模型。而不是指定一个离散序列的隐藏层，我们参数化的衍生物的隐藏状态使用神经网络。使用黑盒微分方程求解器计算网络的输出。这些连续深度模型具有恒定的内存成本，使其评估策略适应每个输入，并且可以显式地以数字精度换取速度。我们在连续深度残差网络和连续时间潜变量模型中证明了这些性质。我们还构建了连续的规范化流程，生成模型，可以训练最大似然，而不分区或排序的数据维度。为了训练，我们展示了如何通过任何ODE求解器可伸缩地反向传播，而无需访问其内部操作。这允许在更大的模型中进行ODE的端到端训练。
英文摘要：
We introduce a new family of deep neural network models. Instead of specifying a discrete sequence of hidden layers, we parameterize the derivative of the hidden state using a neural network. The output of the network is computed using a black-box differential equation solver. These continuous-depth models have constant memory cost, adapt their evaluation strategy to each input, and can explicitly trade numerical precision for speed. We demonstrate these properties in continuous-depth residual networks and continuous-time latent variable models. We also construct continuous normalizing flows, a generative model that can train by maximum likelihood, without partitioning or ordering the data dimensions. For training, we show how to scalably backpropagate through any ODE solver, without access to its internal operations. This allows end-to-end training of ODEs within larger models.
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。
