Deep Signature Transforms

Patric Bonnier, Patrick Kidger, Imanol Perez Arribas, Cristopher Salvi, Terry Lyons
5/21/2019
cs.LGstat.ML

Official code: patrick-kidger/Deep-Signature-Transforms (98 stars, Apache-2.0, last commit 2024-07-25).

Abstract

The signature is an infinite graded sequence of statistics known to characterise a stream of data up to a negligible equivalence class. It is a transform which has previously been treated as a fixed feature transformation, on top of which a model may be built. We propose a novel approach which combines the advantages of the signature transform with modern deep learning frameworks. By learning an augmentation of the stream prior to the signature transform, the terms of the signature may be selected in a data-dependent way. More generally, we describe how the signature transform may be used as a layer anywhere within a neural network. In this context it may be interpreted as a pooling operation. We present the results of empirical experiments to back up the theoretical justification. Code available at https://github.com/patrick-kidger/Deep-Signature-Transforms.

View on arXivView PDF

Code Implementations(1)

patrick-kidger/Deep-Signature-TransformsOfficial100%
Apache-2.0

Code for "Deep Signature Transforms" (NeurIPS 2019)

9821Python, Jupyter NotebookMay 21, 20192 years agoApache-2.0
machine-learningrough-pathssignaturesignaturestime-series

Similar Papers

A Primer on the Signature Method in Machine Learning

84% match1 repo3132016

Learning with Expected Signatures: Theory and Applications

83% match2025

Lecture notes on rough paths and applications to machine learning

82% match2024

Path Signatures for Feature Extraction. An Introduction to the Mathematics Underpinning an Efficient Machine Learning Technique

81% match2025

Online Signature Verification Using Augmented Path Signature and T-Mamba

80% match1 repo12026

Scalable Machine Learning Algorithms using Path Signatures

80% match2025

Cite this paper

@article{bonnier2019deep,
  title  = {Deep Signature Transforms},
  author = {Patric Bonnier and Patrick Kidger and Imanol Perez Arribas and Cristopher Salvi and Terry Lyons},
  year   = {2019},
  eprint = {1905.08494},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1905.08494}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1905.08494.svg)](https://codeofpaper.com/paper/1905.08494)

Discussion