Signature moments to characterize laws of stochastic processes
arXiv:1810.10971
Abstract
The sequence of moments of a vector-valued random variable can characterize its law. We study the analogous problem for path-valued random variables, that is stochastic processes, by using so-called robust signature moments. This allows us to derive a metric of maximum mean discrepancy type for laws of stochastic processes and study the topology it induces on the space of laws of stochastic processes. This metric can be kernelized using the signature kernel which allows to efficiently compute it. As an application, we provide a non-parametric two-sample hypothesis test for laws of stochastic processes.
42 pages. Restructured the text, changed experiments in final section. To appear in Journal of Machine Learning Research
References in corpus (1)
Cited by in corpus (18)
- Neural Controlled Differential Equations for Irregular Time Series
- Persistence paths and signature features in topological data analysis
- Deep Signature Transforms
- Conditional Sig-Wasserstein GANs for Time Series Generation
- A Kernel Two-Sample Test for Functional Data
- Metrizing Weak Convergence with Maximum Mean Discrepancies
- Embedding and learning with signatures
- Signature Cumulants, Ordered Partitions, and Independence of Stochastic Processes
- General Signature Kernels
- Higher Order Kernel Mean Embeddings to Capture Filtrations of Stochastic Processes
- Distribution Regression for Sequential Data
- Feature Engineering with Regularity Structures
- Signatured Deep Fictitious Play for Mean Field Games with Common Noise
- Statistical Depth Meets Machine Learning: Kernel Mean Embeddings and Depth in Functional Data Analysis
- Path Imputation Strategies for Signature Models of Irregular Time Series
- Path Signatures on Lie Groups
- A Fourier representation of kernel Stein discrepancy with application to Goodness-of-Fit tests for measures on infinite dimensional Hilbert spaces
- Deep Signature FBSDE Algorithm