Good explanation for the initial steps of organizing the ML project and the direction to approach the problem accounted for. The quiz was interesting but as it is the same set of questions for any next attempt, I would not say I have mastered the course completely.

Only 2 weeks, good concepts to know. But videos are not "final release" they are not well edited. Some time Andrew repeat the same sentence 2x but they forgot to cut it.

No programming assignment. Although quiz format is fun and you really learn by doing the quiz.

This was a slightly more theoretical course than the first 3 in the Deep Learning specialization and, even thought I enjoyed it, I think the info would stick better if there would have been a programming assignment too (or some other type fo practical application).

Nice and concrete examples of what to think of and focus on when trying to improve your machine learning projects. Not as engaging tasks to complete as in the previous courses in this specialisation, however a good change of scenary if you have been doing these.

Great material, but it's not quite easy to understand it from scratch, if you didn't have such problems yourself (i.e if you have no experience in deep NN training). I've stored this material and going to revisit it after I gain more experience in training NNs.

Great course, however the quiz of week 2 had some ambigious phrasings and I think at least one example (the one with the data synthesis of foggy images) is contradictive of what was taught in the video lessons. Other than that, really good content and teaching!

A lot of concepts were put forward and taught well. If there was a programming assignment as well to back up the concepts that were taught like multi-task learning, how to deal with data mismatch, dividing the total data into train\train-dev\dev\test data etc.

I feel there is a glitch because even after completing the videos, it is not shown as completed and I had to replay them multiple times. Also there is a glitch in the assignment, because the correct answer in one attempt is shown as incorrect in the next try

Great and easy-to-follow introduction to structuring machine learning projects and focusing on what to tune on neural networks. One star left out because the content has not been polished, but there are minor errors here and there with separate corrections.

The course seems a little less concrete than the others in this specialisation. But nevertheless, still a useful building block in anyone's deep learning repertoire. And note it will probably take less time to complete than the others, so plan accordingly.

It would be nice if there are hands on assignment or small projects on fine-tuning with existing weight you can found in the internet or multi-task learning project. Overall, it's a great course with many useful technique to try in the real world projects.

The course was really great, but a little part of the content was repeated from previous courses of the specialization. Also there should be more quizzes or exercises to master the large amount of practical advices for managing machine learning projects.

Very good class! It really hits me that AI programming is a different paradigm. Managing data is key.

That said, the materials in week1 have quite a bit of overlap with the 1st course (NN & DL). The materials should simply do a quick reminder a move on.

I liked this course, but not as much as the others. It is however setting the foundation for the remainder of the course material. It carries with it wisdom, which I think will make more sense at a later point when confronted with real life challenges

This is the course that really confirms Andrew Ng's grasp on the practically application of AI and ML. As long as you pay attention to what is said, you will get a lot from this course. I wish there was an edited collection of notes for this course

Very useful part of the course set. Would like it to be slightly longer with more examples of TensorFlow and Keras. I felt that I put to much effort of trying to understand Keras in course number 5 instead of learning the principles and algorithms.

I like the "flight simulator" quizzes a lot and other courses might benefit from a similar assessment (in addition to regular quizzes and programming exercises), but I do think this course would benefit from some programming exercises too. Thanks!

It is a great course! thanks everyone involved in making it! If you can make more questions in every video or so i think it would be better. Also different case scenarios like the one you presented here in order for students to generalize better

I enjoyed the course content a lot, but noticed a lot of errors in test materials and test sections that didn't seem to make sense. For example, there were references to a flight simulator in quizzes that actually never appeared in any questions

I wish it could have had some coding practices and more mathematical insights. For instance, more insights on the different metrics (precision, recall, f-measure) and having some coding practices to get better sense of it on real world examples!

Very useful and practical information. Some of the videos had sound issues and minor editing glitches. I wish there had been more hands-on assignments. The few that were supplied were good but I think one more, less guided example would be good.

Comparing other courses before, it focus more on structing deep learning program and evaluate properly. However, the content in this week is really boring. In my opinion, it is better to imptove the course by teaching more implementation codes.

Some questions in week 2 test are ambiguous and the last videos have edition errors. But overall strategies for guiding projects are very useful. It's a great course about practical aspects of Deep Learning you'll probably not find anywhere.

I have previously worked on DL project and its performance was not good in real-world data, I wasn't able to draw any reason for it. This course taught me how to deal with such kind of problem and how one can approach the possible solution.

Was a good course with a lot of useful tips that I am sure I am able to use in my job as a data scientist. However, I would've liked if there were a few more hands-on examples (e.g. using jupyter) to really drives these concepts more home.