Really bad best practices on splitting data. You should use random functions with a seed instead. The question on whether to use the first few digits or last few digits of a hashing function is totally arbitrary. It doesn't matter. You shouldn't use either. You should use ALL the digits of the hashing function. Also, the code shown doesn't look at the last digits, it uses modulo. Which is not the last few digits unless your modulus is a power of ten.

It's unclear for who this course is meant. It mixes basics like train-validate-test with lectures that use machine learning terms that only have meaning to someone who has already knows ML terminology. If you're looking for a good introduction to ML - check out Andrew Ng's course.

"Short History of ML" was good, if kind of light. The rest of this course is flaming garbage. Zero topic organization. Material is poorly explained. Labs are poorly detailed and in some cases don't work out of the box. Seriously, skip this course and take Andrew Ng's instead.

The notebook based demos are unfortunately pretty useless as labs. All of these courses would be much improved with real labs that require the student to build the system.

Pros: Tensorflow is an excellent framework for deep learning

Cons :

1- The way this material is designed is 10 X SHIT

2- Either teach properly or don't teach at all.

I'm not able to access the video lectures even though I purchased and completed this course 6 months back

Too much scary information provided at once combined with the mindbending sql queries and data parsing

This is the most superficial course I have ever taken. I just waisted my time.

Cannot access resources after completing the course

Did not learn much! Poorly instructed courses.

not well structured and lack handouts

No hands on learning.