This is such a well-written article. However, I do have a few questions. Why use regressor but not classifier when defining the ‘model_factory’? I am fairly new to semi-supervised learning and have just begun studying about said topic. It would be very helpful if someone could give an answer to my question.
Link to the article: