Refactor and merge all code into master

[miguelgfierro]

• [x] Decide the experiments we want to present
• [x] Clean code and make comments
• [x] Merge code in master

[miguelgfierro]

What we have: -lightgbm vs xgboost 1) airline dataset 2) BCI 3) football 4) kaggle 5) fraud (new dataset)

-concept drift 1) airline 2) BCI 3) football spanish league

-exclude 1) IoT

[miguelgfierro]

for notebooks:

• lightgbm vs xgboost: result notebook with plots, results and discussion
• concept drift: result notebook (what is concept drift, results of all datasets and discussion)
• common: one notebook for each experiment (data loading, data management, training, cv, print the results (speed+acc+precision+recall+auc), concept drift)

[miguelgfierro]

Create the single notebooks with the complete pipeline