from my perspective, maybe, it would be better if this course is the end course of the specialization. the contents are greate. I would like to suggest others to put this course in the end of the specialization.

Very interesting, but too short. The aim of the course is to provide a good overview of the different situations occuring in a project, but there is more questions arising. Experience will come with training.

The information is very useful, but the text, quizzes, and video all felt a little less polished than the previous two courses. Not a big deal, but it was notable and occasionally distracted from learning.

nice to follow, but I would have liked it there would have been more variance. e.g. quizzes breaking the videos. I'm basically comparing this experiment with the other courses made by Andrew/deeplearning.ai

A useful few hours of videos. I found the questions quite useful, but overall feel this project would have been better off being spread across other weeks, as it doesnt work so well as a stand-alone course.

Very practical. programing assignment using the concepts would help to solidify the concepts. I would really appreciate programming assignments on Transfer learning since a lot of industries practices it.

Clear course, it would have been helpful to add notebooks to the course to have a more realistic feeling of the problems. This would make it also more clear how the dev set influences the training phase.

The course appraises the reader of the various tricks that are needed to design nice machine learning projects. One minor suggestion would be to have some programming assignments for this course as well!

Great material but a few glitches in the videos that need editing out. I'm sure the people responsible will put it right soon though; the course hasn't been out long. Part of an excellent specialisation.

Overall the course was good, but lacked some minor information needed for the quizzes. The quizzes also included some questions that felt like they were meant to trick us rather than test our knowledge.

Lots of good information, and it's quick, but the course videos could use a bit of editing compared to the other ones. It's common to have repeated takes in the videos that should have been edited out.

I didn't find this course as useful as the previous two. I'd still recommend the course because some interesting concepts are covered. The materials seem more intuitive. The quizzes are good practices!

The course is very teaching in my uneducated opinion and will help m later in life, hopefully.

I wish the test question had been more coherent.

I enjoyed learning it, and the simulator is a great idea!

Course is great although only in one case we have pdf's with additional lecture notes. They are more useful than ppt slides therefore it would be desirable to have them in other parts of the course

The course is perfect. But it would have been better if some ways implementation was also introduced to us. But overall Andrew is really helpful to make us visualize and understand new strategies.

In this section, only strategies discussed and no programming practice. So I don't think I really understand those tricks. Maybe when I enroll in real projects, those magic will show their powers.

Very good lessons to teach the principle of how to set up a machine learning project. Lots of tricks are taught, but still hoping there could be some practical assignment to enhance the knowledge.

Videos could be shorter and better edited. Supporting materials could be better. Homework assignments were tedious and included rather confusing (and borderline tricky) questions and alternatives.

It is a useful Content to keep in mind. Not as practical as other contents in the Specialisation. But, as Andrew Ng pointed out, it is a topic many times overlooked and I'm glad it was discussed.

The course is the best available on the online education platforms so far. An excellent instructor and really engaging assignments. You get it all but it lacks availability or reading materials.

Lot of things to intake , Error analysis seemed more difficult to me compared to Bias/Variance Analysis. I think lot of practice is required to really remember and apply in real world problems.

There are some answers on the "flight simulator" that are ambiguously worded, and one that seems to flat-out contradict what Prof. Ng said in lecture. Search the discussion forums for "foggy".

Loved the course, and the simulation was great. Doing an actual transfer learning programming exercise in TensorFlow could be an awesome addition.

Best & thanks again for an awesome course!

Eric

Basic but super relevant and well structure! Things we may all know, but forget to think about... I highly recommand. The only reason why I did not give 5 stars is that it's sometime repetitive

Course was really interresting with lot of insights. I would have maybe put even more examples/cases and added more quizz to get more practical training if we apply the pilot training analogy.