Chris White
Chris White
All good questions; the current solvers which are immediately generalizable to other models are: `gradient_descent`, `admm`, and `newton` by replacing function / gradient / hessian evaluations with calls to the...
Could you provide an example of a feature we gain by inheriting directly from `sklearn` ? I don't clearly see what these `sklearn` classes provide us.
Still need abstract base class for families.
Yea that sounds like a good plan; I can look into this more on Thursday / Friday.
See #12
For unregularized problems, we can test for optimality of arbitrary problems easily by using `y = sigmoid(X.dot(beta))` (no thresholding) and testing that the estimated coefficients are close to `beta`; how...
Some examples of the issues we face in implementing a robust testing framework can be found here: https://github.com/dask/dask-glm/blob/master/notebooks/AccuracyBook.ipynb
https://gist.github.com/moody-marlin/e2de54ca17d615b263f80372031cb865 cc: @mpancia Proximal grad does worst because the line search is currently very crude.
@mrocklin Has `gradient_descent` been optimized (using delayed, persist, etc.) in the same way that the other functions have? I might be refactoring soon and I wanted to make sure that...
@mrocklin Did you look at ADMM? I'm currently starting to think that, going forward, we only employ ADMM, Newton, and gradient_descent.