[ENH] Riemannian distance and Riemannian classifiers

[TonyBagnall]

Describe the feature or idea you want to propose

Riemannian distance and Riemannian classifiers are popular for EEG classification in BCI field. It would be interesting to compare them to tsc algorithms.

Describe your proposed solution

we could introduce a dependency for this
https://github.com/pyRiemann

or implement it from scratch as an aeon distance function

[TonyBagnall]

on futher reading of https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5444491

The Reimannian distance is the distance between the power spectral density matrix which is found with the the Nuttall-Strand algorithm [18], [20]

"The algorithm uses forward-backward linear prediction to iteratively estimate the residual covariance matrix arriving at an accurate positive semi-definite estimate of the power spectral density matrix with high frequency resolution. This estimate will be used as the power spectral density feature to characterize the EEG signals"

https://ieeexplore.ieee.org/document/6319854