Hello, I will try not to give the answer.
In question 2, it gave me an error, when I use model.get_params() for logisticregression I got the answer… the answer in the question is for the histgradientboostingclassifier model.
Hello! I am not sure if I understand correctly but please take into account that when building a pipeline using make_pipeline
, it will give standardized names to each step. If you build it using Pipeline
, then you have to provide aliases to each step, for instance:
pipeline = Pipeline([
('scaler', StandardScaler()),
('classifier', LogisticRegression())
])
means that we are asigning the alias 'classifier'
to the logistic regression step and therefore, any hyperparameter would follow the notation alias__hyperparameter
.
This is implicetly shown in the video Analysis of hyperparameter search results.