Thanks for explaining that. It seems that penalty is already used, which is unfortunate. One of the interesting things about ML is that you can learn for a year

Thanks for explaining that. It seems that penalty is already used, which is unfortunate. One of the interesting things about ML is that you can learn for a year and a half and still uncover more things you didn't know, which I love. I guess I'll call loss "punishment." It matches how it feels to make progress in ML anyway.