[ENH] AutoETS implementation

[TonyBagnall]

Describe the feature or idea you want to propose

Our new experimental forecasting module starts with a really fast ETS implementation. The next stage is AutoETS. But how to search the parameter space? There are many alternatives, grid search, nelder mead, stochastic gradient descent etc. It would be really good to implement this in a configurable way. Feel free to come up with alternative heuristic search algorithms on this thread

Describe your proposed solution

we need to think about how to design this efficiently, perhaps taking inspiration from scikit-learn?

Describe alternatives you've considered, if relevant

No response

Additional context

No response

[TonyBagnall]

@aeon-actions-bot assign @alexbanwell1

[SebastianSchmidl]

Bayesian optimization using https://optuna.readthedocs.io/en/stable/?