Add number_samples to XGBoost ML Models

[V1NAY8]

Work on #243

TreeNode of xgboost has the following information:

Tree               1
Node               0
ID               1-0
Feature           f1
Split      -0.030468
Yes              1-1
No               1-2
Missing          1-1
Gain       44.768673
Cover      23.395073
Name: 7, dtype: object

So, based on

• importance_type parameter of https://xgboost.readthedocs.io/en/latest/python/python_api.html#xgboost.XGBRegressor
• based on https://datascience.stackexchange.com/a/12594

I am thinking Cover gives us the number of training samples that is being covered by the node or leaf since "Gain" is already used as leaf_value. I might be wrong though!

@benwtrent Please give a review :)

[elasticmachine]

Since this is a community submitted pull request, a Jenkins build has not been kicked off automatically. Can an Elastic organization member please verify the contents of this patch and then kick off a build manually?

[benwtrent]

I don't think this is gonna work.

cover is a float. I don't know how they calculate cover related to the total number of samples that hit the tree. The docs indicate its the average of the total number of samples effected by the split.

I wonder if cover * number of features ~= The total number of docs? Can you investigate and confirm?