[ENH] Reorganise the structure of the clustering module

[TonyBagnall]

Describe the feature or idea you want to propose

currently contains the following submodules

feature_based
deep_learning
averaging
compose

Describe your proposed solution

averaging is to going to move to protoyping

after meeting on 25/10/24 propose adding

1. hierarchical (when we get a hierarchical algorithm)
2. partional for all of the k-means and medoids algorithms

Describe alternatives you've considered, if relevant

No response

Additional context

No response

[SebastianSchmidl]

Is there a roadmap for adding hierarchical clustering algorithms?

I have some code lying around that uses sklearn.cluster.AgglomerativeClustering with metric="precomputed" and the aeon distances. It was rather straight-forward until I wanted to use ward-linkage with any elastic distance because sklearn forbids using ward-linkage with anything other than Euclidean distance. I know it is not supposed to be used like that but provides good results.

[TonyBagnall]

No roadmap, I think @chrisholder has some kicking around t go in soon.