DeepAI AI Chat
Log In Sign Up

Path Imputation Strategies for Signature Models

by   Michael Moor, et al.
ETH Zurich

The signature transform is a 'universal nonlinearity' on the space of continuous vector-valued paths, and has received attention for use in machine learning. However real-world temporal data is typically discretised, and must first be transformed into a continuous path before signature techniques can be applied. We characterise this as an imputation problem, and empirically assess the impact of various imputation techniques when applying signatures to irregular time series data. In our experiments, we find that the choice of imputation drastically affects shallow signature models, whereas deeper architectures are more robust. We also observe that uncertainty-aware predictions are overall beneficial, even compared to the uncertainty-aware training of Gaussian process (GP) adapters. Hence, we propose an extension of GP adapters by integrating uncertainty to the prediction step. This leads to competitive performance in general, and improves robustness in signature models in particular.


page 4

page 8


Tropical time series, iterated-sums signatures and quasisymmetric functions

Driven by the need for principled extraction of features from time serie...

Mixture-based Multiple Imputation Models for Clinical Data with a Temporal Dimension

The problem of missing values in multivariable time series is a key chal...

On Consistency of Signatures Using Lasso

Signature transforms are iterated path integrals of continuous and discr...

LORD: Lower-Dimensional Embedding of Log-Signature in Neural Rough Differential Equations

The problem of processing very long time-series data (e.g., a length of ...

Distribution Regression for Continuous-Time Processes via the Expected Signature

We introduce a learning framework to infer macroscopic properties of an ...

Towards a more efficient representation of imputation operators in TPOT

Automated Machine Learning encompasses a set of meta-algorithms intended...

A Bayesian Decision Support System in Energy Systems Planning

Gaussian Process (GP) emulators are widely used to approximate complex c...

Code Repositories