We introduce three new generative models for time series. Based on Euler discretization and Wasserstein metrics, they are able to capture time marginal distributions and temporal dynamics. Two of these methods rely on the adaptation of generative adversarial networks (GANs) to time series. Both of them outperform state-of-the-art benchmarks by capturing the underlying temporal structure on synthetic time series. The third algorithm, called Conditional Euler Generator (CEGEN), minimizes a dedicated distance between the transition probability distributions over all time steps. In the context of Ito processes, we provide theoretical guarantees that minimizing this criterion implies accurate estimations of the drift and volatility parameters. We demonstrate empirically that CEGEN outperforms state-of-the-art and GAN generators on both marginal and temporal dynamics metrics. Besides, it identifies accurate correlation structures in high dimension. When few data points are available, we verify the effectiveness of CEGEN, when combined with transfer learning methods on Monte Carlo simulations. Finally, we illustrate the robustness of our method on various real-world datasets.
翻译:我们引入了三种新的时间序列遗传模型。 根据Euler离散和瓦塞斯坦度量, 它们能够捕捉时间边际分布和时间动态。 其中两种方法依靠基因对抗网络(GANs)适应时间序列。 这两种方法都通过在合成时间序列中捕捉基本时间结构而优于最先进的基准。 第三个算法称为“ 有条件电动生成器(CEGEN), 最大限度地缩小所有步骤的过渡概率分布之间的专门距离。 在Ito 进程中, 我们提供理论保证, 最大限度地减少这一标准意味着精确估计漂移和波动参数。 我们从经验上证明, CEGEN 超越了边际和时间动态指标上的最新和GAN 生成器。 此外, 它确定了高维度的准确关联结构。 当数据点很少时, 我们核查CEGEN的有效性, 与蒙特卡洛模拟的传输学习方法相结合。 最后, 我们用各种现实世界数据集展示了我们的方法的稳健性。