cnellington / Contextualized

An SKLearn-style toolbox for estimating and analyzing models, distributions, and functions with context-specific parameters.
http://contextualized.ml/
GNU General Public License v3.0
65 stars 9 forks source link

Refactored networks.py into networks module #78

Closed cnellington closed 2 years ago

cnellington commented 2 years ago

Resolves #63

cnellington commented 2 years ago

Correlation is a specific use case of univariate regression, so I agree it makes more sense to have a general correlation class / trainer in the regression module. I'll add what we have now to regression and make an issue to add the general ContextualizedCorrelation(C, X, Y) in the future.

On that note, I'll organize the contextualized package into model-class modules so that we have minimal overlap between metamodels, optimizers, and losses.