Dynamic Time Worping and Automal Transport Methods

The grumovasterstein distance is essentially the kind of interrelational geometries, as we call them, between the two sets of samples. Different points in this data set have a different number of discreet observations becausein some of them they're drawn faster than in other ones. And so one of the things that methods like dynamic time worping and actually automal transport methods also allow you to do is they allow you to have notions of berry centrs. Now a berry centre is like a mean, but a mean is something that you have of real numbers,. whereas a berry centere can be something you can have of a something much more complex.