Analysis class needs to be beefed up with something actually useful

[ardunn]

Given a PredictionPipeline object (or just a tpot model and feature list until I get the top level classes working better), analysis should give back a nice html (or other format?) containing:

• identification of outliers
• identification of the most important features
• partial dependence plots w/ skater
• LIME plots w/ skater
• details of the features dropped/features retained
• breakdown of the featurization time/fitting time/etc.
• t-SNE plot based on features and labelled by material formula/phase

@Doppe1g4nger @ADA110 any other ideas on cool things to include here?

[albalu]

For outliers, there is single class SVM and isolation forest which are basically what they sound like but with one-class classification. In isolation forest you run random forest and the points that are identified by too few splits can be the outliers. This could be used as a default. I wanted to implement it but don't think will have time :| It is already implemented in sklearn but someone should just integrate it in the workflow

[Doppe1g4nger]

Also we should add a tab of data giving a description of how the model that was selected works, at least in a shallow not-directed-towards-experts manner.