Gaussian processes based data augmentation and expected signature for time series classification

2023/10/16 by Romito, Marco, Triggiano, Francesco
#FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2310.10836

Abstract

The signature is a fundamental object that describes paths (that is, continuous functions from an interval to a Euclidean space). Likewise, the expected signature provides a statistical description of the law of stochastic processes. We propose a feature extraction model for time series built upon the expected signature. This is computed through a Gaussian processes based data augmentation. One of the main features is that an optimal feature extraction is learnt through the supervised task that uses the model.

Related

• A Generalised Signature Method for Multivariate Time Series Feature Extraction
• Bayesian Learning from Sequential Data using Gaussian Processes with\n Signature Covariances
• Learning with Expected Signatures: Theory and Applications
• Learning to Forget: Bayesian Time Series Forecasting using Recurrent Sparse Spectrum Signature Gaussian Processes
• Path Imputation Strategies for Signature Models of Irregular Time Series