It would be helpful if more checkpoints/unit tests were added to the Jupyter notebooks.
This post to the Classification course's discussion forums specifically requests more checkpoints for the "Implementing logistic regression from scratch" notebook. I think that this is a good idea because there is at least one thread about this particular notebook in each session.
Another notebook that students have trouble with is "Boosting a decision stump". In particular, the calculation of weight and weighted_error is often incorrect.
It would be helpful if more checkpoints/unit tests were added to the Jupyter notebooks.
This post to the Classification course's discussion forums specifically requests more checkpoints for the "Implementing logistic regression from scratch" notebook. I think that this is a good idea because there is at least one thread about this particular notebook in each session.
Another notebook that students have trouble with is "Boosting a decision stump". In particular, the calculation of
weightandweighted_erroris often incorrect.