Articles by David Hirnschall
The log-signature-based time series Wasserstein generative adversarial network
This paper puts forward a method using neural networks to minimize the Wasserstein distance between log signatures to reduce the target dimension of the generator and increase performance