Adeline Fermanian
A statistical point of view on signatures
Abstract
We will be concerned with the application of signatures to machine learning and statistics. The basic principle of the signature method is to represent multidimensional paths by a graded feature set of their iterated integrals, called the signature. After a general overview of signatures in machine learning, we will focus on two specific problems: estimation in a linear regression setting and building a canonical signature pipeline for time series classification.
