Add plotting of result histograms in TMVACrossValidationApplication tutorial

* Update GenerateModel.py

keras.utils.visualize_util is outdated

* Update GenerateModel.py

They were missing a crucial comment header, making the documentation
generation system not pick them up.

* [TMVA] CV - Add workaround for large event numbers

Currently TMVA considers all internal variables floats, if large event
numbers are then input as part of "cv-in-app" these can overflow the
float resulting in erroneous fold assignments.

A workaround is to reduce the input range when reading the input from
the TTree as TTreeFormula retains awareness of the input type.

[TMVA] Configure tutorials/tmva/makefile

Example makefile from tutorials/tmva should use the same c++ standard as
ROOT (otherwise one faces errors from string views in the root headers
when building the tutorials).

Based on the solution exposed here: https://github.com/root-project/root/commit/9339de9e644b06d5f77607f83aea118969cab7b6

Removes unused spectator expression.
Removes weird weighting and replaces it with an informative comment.

* [TMVA] CV - fix not able to compile with cling

* [TMVA] CV - Rename variable `ce` -> `cv`

* [TMVA] CV - fix out of range crash

Method signature used for `DataLoader::PrepareTrainingAndTestSet`
triggered the generation of classes "Signal" and "Background". Since the
task was regression this caused confusion further down the processing
pipeline.

Code changed to use global selection cut instead of implicit
signal/background cuts.

* [TMVA] CV - improve documentation

* [TMVA] improve documentation of TransformationHandler::AddStats

Speed up  TMVA regression tutorial by switching off by default the old MLP and GA methods. Add instead DNN when TMVA supports CPU.

- Variable renaming ce -> cv.
- Remove extraneous text output from method logs.

[TMVA] Remove unused helper scripts which create errors in the incremental tests; update Keras tutorials to 2.x API

removed 2 empty lines

batchSize

default

clang fix