Preprocessing for numerical features

Module 1. The Predictive Modeling Pipeline Fitting a scikit-learn model on numerical data
#1

My comments :

  • “instead to read” => “instead of reading” (first block)

  • By contrast, with the previous notebook, we will select the numerical features from the original dataset instead to read an external file where the selection has been done beforehand. I think that we did the exact same thing in the previous notebook already. Not sure this statement is useful here.

  • it could be interesting to compare the statistical performance of the logistic regression computed with cross-validation performance of the predictor used in the first notebook.

#2

I think I have mostly tackled these points.

#3