Should we include the regulariser penalty into the cost function ?
Hello,
I have a question about the grading scheme for the projet 1.
One of the remarks about our code was that we did not include the regularisation penalty term in cost functions.
However, during the lab, to have a loss curve that match the one given in the lab handout, we saw that we shouldn't include the regularisation penalty term. That's why we did not include it in our project. Furthermore, if we look at the lab corrections, they do not include the regularisation penalty term. Since we should have implemented the project with the functions made during the lab, we just want to understand which approach is correct in practice and for the rest of the course, should we include the regularisation penalty term or not in our loss function ?
Should we include the regulariser penalty into the cost function ?
Hello,
I have a question about the grading scheme for the projet 1.
One of the remarks about our code was that we did not include the regularisation penalty term in cost functions.
However, during the lab, to have a loss curve that match the one given in the lab handout, we saw that we shouldn't include the regularisation penalty term. That's why we did not include it in our project. Furthermore, if we look at the lab corrections, they do not include the regularisation penalty term. Since we should have implemented the project with the functions made during the lab, we just want to understand which approach is correct in practice and for the rest of the course, should we include the regularisation penalty term or not in our loss function ?
Thanks a lot for your answer !
Alessio
Add comment