Add Nelder-Mead simplex about optimize

[UserAB1236872]

I just came here to add this. I'm looking at the Sci-Py code right now and can start working on this soon.

[UserAB1236872]

https://github.com/scipy/scipy/blob/master/scipy/optimize/optimize.py

_minimize_neldermead if anyone is curious. (I have the paper handy too, but real code can be nicer)

[btracey]

When you're doing that, please feel free to add documentation / add issues for specific pieces of documentation. We definitely need a better doc comment for using optimize, but if it's confusing how to create a new optimizer even knowing how to use the package, we should work to improve that too.

[UserAB1236872]

Maybe a separate issue, but as an idea maybe putting some well documented ExampleX_ in the tests might help. Something even simpler than Rosenbrock like maybe batch/stochastic Gradient Descent for SSE (or negative log likelihood for Logistic Regression) with varying degrees of regularization. I think that would be a good walk-through of building up how to do simple things step by step one piece at a time.

[vladimir-ch]

Closed by 47bdee1 in #91