The Signature Kernel
This is an incremental survey paper providing an elementary introduction to the signature kernel for researchers in machine learning and data analysis.
The paper introduces the signature kernel, a positive definite kernel for sequential data, which offers theoretical guarantees from stochastic analysis, efficient computational algorithms, and strong empirical performance.
The signature kernel is a positive definite kernel for sequential data. It inherits theoretical guarantees from stochastic analysis, has efficient algorithms for computation, and shows strong empirical performance. In this short survey paper for a forthcoming Springer handbook, we give an elementary introduction to the signature kernel and highlight these theoretical and computational properties.