Ensemble based on boosting

Hi, I did not follow the example of correct prediction:


the numbers do not match ?! Could you explain please ?

Yes, indeed we changed the dataset somehow and the explanation does not fit anymore. I will push a fix. However, you can change the cell [7] to pick a sample that make sense with the explanation:

data_max = data_train.iloc[-2, 0]
target_true = target_train.iloc[-2]
target_true_residual = residuals.iloc[-2]

Solved in FIX select a sample matching explanation · INRIA/scikit-learn-mooc@3fc6934 · GitHub

One needs to synchronize the notebooks in FUN