BaseEstimator and Training a dataset

  1. Sir, I want to know basically what does the BaseEstimator and TransformerMixin do in the above class if we pass the values like age, SibSp, parch, fare. What is the function of BaseEstimator and TransformerMixin and why do we use that.
  2. I have another doubt, Let us say if we are training a dataset using SGDClassifier we are getting output with an accuracy of 95%. So why don’t we again train the trained dataset using some other classifier like RandomForest so that our model’s output accuracy might be much higher…???

In sklearn, when you have to create your own custom pipeline, you need to extend BaseEstimator and TransformerMixin.