Pinned Repositories
esig
esig python package
libalgebra-lite
nlpsig
Package for constructing paths of embeddings obtained from transformers.
PyRecombine
pysegments
RoughPy
Toolbox for working with streaming data as rough paths in Python
signatory
Differentiable computations of the signature and logsignature transforms, on both CPU and GPU. (ICLR 2021)
signature_applications
Example applications of path signatures
signature_mahalanobis_knn
Methodology for anomaly detection on multivariate streams using path signatures and the variance norm.
SigNova
DataSig's Repositories
datasig-ac-uk/esig
esig python package
datasig-ac-uk/signature_applications
Example applications of path signatures
datasig-ac-uk/RoughPy
Toolbox for working with streaming data as rough paths in Python
datasig-ac-uk/signature_mahalanobis_knn
Methodology for anomaly detection on multivariate streams using path signatures and the variance norm.
datasig-ac-uk/nlpsig
Package for constructing paths of embeddings obtained from transformers.
datasig-ac-uk/pysegments
datasig-ac-uk/SigNova
datasig-ac-uk/signatory
Differentiable computations of the signature and logsignature transforms, on both CPU and GPU. (ICLR 2021)
datasig-ac-uk/libalgebra-lite
datasig-ac-uk/path_signatures_introduction
A Brief Introduction to Path Signatures for Machine Learning Practitioners
datasig-ac-uk/PyRecombine
datasig-ac-uk/Recombine
A shared C++ library for dynamic Caratheodory data simplification - has a robust and stable C interface
datasig-ac-uk/DatasetLoader
A utility project to provide a convenient and consistent access to various datasets.
datasig-ac-uk/diffrax
Numerical differential equation solvers in JAX. Autodifferentiable and GPU-capable. https://docs.kidger.site/diffrax/
datasig-ac-uk/PSFDataset
Implementation of the Path-signature-featuremethodology for creating datasets for human action recognition from landmark data.
datasig-ac-uk/Sepsis_Label_Extraction
datasig-ac-uk/torchcde
Differentiable controlled differential equation solvers for PyTorch with GPU support and memory-efficient adjoint backpropagation.