Scalable Signature Kernel Computations for Long Time Series via Local Neumann Series Expansions

TL;DR

This paper introduces an efficient, scalable method for computing the signature kernel of long, high-dimensional time series using local Neumann series expansions, enabling high accuracy and reduced memory usage.

Contribution

The paper presents a novel adaptive local power series expansion technique for the signature kernel, improving computational efficiency and scalability for very long and high-dimensional time series.

Findings

Achieves substantial performance improvements over existing methods.

Provides adjustable accuracy suitable for high-roughness data.

Enables handling of very long time series (over one million points) on a single GPU.

Abstract

The signature kernel is a recent state-of-the-art tool for analyzing high-dimensional sequential data, valued for its theoretical guarantees and strong empirical performance. In this paper, we present a novel method for efficiently computing the signature kernel of long, high-dimensional time series via adaptively truncated recursive local power series expansions. Building on the characterization of the signature kernel as the solution of a Goursat PDE, our approach employs tilewise Neumann-series expansions to derive rapidly converging power series approximations of the signature kernel that are locally defined on subdomains and propagated iteratively across the entire domain of the Goursat solution by exploiting the geometry of the time series. Algorithmically, this involves solving a system of interdependent Goursat PDEs via adaptively truncated local power series expansions and recursive propagation of boundary conditions along a directed graph in a topological ordering. This method strikes an effective balance between computational cost and accuracy, achieving substantial performance improvements over state-of-the-art approaches for computing the signature kernel. It offers (a) adjustable and superior accuracy, even for time series with very high roughness; (b) drastically reduced memory requirements; and (c) scalability to efficiently handle very long time series (one million data points or more) on a single GPU. As demonstrated in our benchmarks, these advantages make our method particularly well-suited for rough-path-assisted machine learning, financial modeling, and signal processing applications involving very long and highly volatile sequential data.