Going beyond the technical details, this part of the course goes into the high level view on how to direct your efforts in a ML project. Really enjoyable and useful. Thanks for making this available!

This course presented repeated some of the material from previous courses, had limited challenging material and no programming. It was much too easy for anyone who had already completed the first two courses and it should not have been a standalone course but rather could easily have been part of another course.

Time wasting, all could be summarized in 30 mins video at the end of the previous course

This specialization has increased my knowledge and passion to learn about machine learning.

but that course took me alot as i really hated wasting my time watching aaaaalllll these videos for nothing really really small amount of useful information

Sorry if i was rude, but that's my opinion and that's because i really appreciate coursera contribution in knowledge sharing especially for those who can't afford it (like me)

I just felt this course was a bit confusing in the sense that I never really got the chance to apply what I have learned all in this course. For example, for transfer learning, it's much better to have a programming assignment to understand and test as to for what kind of problems is this gonna be necessary and where can I go wrong instead of just a set of questions.

This seriously shouldn't have been a 2 weeks course, Instead, this should have been a 4 weeks course to better understand the intricacies in the learning algorithms and it's diversifications.

I was disappointed.

I learned so many things in this module. I learned that how to do error analysys and different kind of the learning techniques. Thanks Professor Andrew Ng to provide such a valuable and updated stuff.

This course would be immensely helpful for those who have not started on their first machine learning project. However, the insights shared are quite commonsensical and intuitive for those who have already had some minimal experience in machine learning. This course also does not feel as substantial as the other courses in the specialization, though the tips provided are definitely valuable.

Though it might not seem imminently useful, the course notes I've referred back to the most come from this class. This course is could be summarized as a machine learning master giving useful advice.

This is the knowledge in which we will get from lots of experience only, but the andrew has shared in this course which might help us in future by saving a lot of time through this course experience

This course is too elementary and abstract compared to previous two courses. It is more for a folks managing DL/ML projects . I would have expected more hands on coding experience for much deeper concepts in this DL course rather some very elementary theoretical discussion on how to Manage ML Projects.

This is a must course in the entire specialization. It covers the step by step procedure to approach and solve a problem. The case studies provided are real world problems which are so much helpful.

I rushed through this course when I was starting with the deep learning thinking this course isn't really that technical. Coming back to it after 2 years of industry experience in deep learning-based algorithm development, I realized out of all the other courses in the specialization, this is the most crucial. This course is the one you will benefit most out of when working on a large-scale project, the when in which strategies are revealed which can be learned only by years of experience.

It is very nice to have a very experienced deep learning practitioner showing you the "magic" of making DNN works. That is usually passed from Professor to graduate student, but is available here now.

To set the context, I have a PhD in Computer Engineering from the University of Texas at Austin. I am a working professional (13+ years), but just getting into the field of ML and AI. Apologies for flashing this preamble for every course that I review on coursera.

This course is the third in the deeplearning.ai series offered by Dr. Andrew Ng. This is a relatively short course as compared to the other courses in this series. However, there are quite a few videos to watch and learn from. This course is really a series of practical advice, strategies and analysis techniques that are an indispensable part of the ML/DL toolbox of a practitioner. The techniques are presented through a series of examples and Dr. Ng helps beat the "practical theory" into the student very well.

I was at first disappointed that there were no programming proects. However, the "flight simulator" quizzes were quite challenging and made me think -- thereby, more than made up for the absense of projects. This course is a critical part of the entire series and it is best understood when taken as a part of the sequence.

Thanks Dr. Ng and teaching assistants. This is a fantastic course. Thanks, coursera.

In my opinion this course is an important part of the specialization. But a number of issues made it fall short of a five star rating.

First the positives:

Of the three courses I've taken so far, the quizzes showed great creativity, which made learning fun. Well done.

Unapologetically this course does not require students to write a single line of code. Instead it necessarily addresses issues that coders and their managers need to consider. It's not 'fluff,' but real decision points for machine learning applications. So I found this course of value for the specialization.

Here's what keeps it from five stars.

I think that there ought to be mini-quizzes at the end of every lecture so as to reinforce learning, rather than once at the end of 10 to 12 videos.

This course is probably the right place in the specialization to include a section on the shortfalls of machine learning. So far the specialization takes the perspective of happy technicians solving the world's problems through machine learning. I think it appropriate to spend some time on how machine learning caused such a grading problem during COVID-19 (in the UK and for the International Baccalaureate), how thousands in state employment can skew results by data input and the like.

Finally, as a small but jarring note - can't somebody re-edit some of the videos? Having audio repeated, hearing "test test," and microphone problems are shortcomings not expected from a course that endeavours to teach professionals.

Overall four stars.

This is something you won't see in every machine learning courses. Well, Course 1 and Course 2 are also good. Andrew definitely has a thing to explain complicated stuff in the easy way, e.g. the part where he explained how Adam works in Course 2 is truly amazing.

But this course is really different. It appears to have no technical details and I can see some people may consider this course worthless or make disparaging remarks. But based on my personal experience, what he taught in this class is really important and kind of shapes the way one needs to think about how to tackle a machine learning project from the start. As researchers and engineers, it's easy for us to delve into the technical details and algorithm approaches, papers too early when the overall direction is not too clear yet. I feel very grateful for what Andrew did by sharing his knowledge to all of us.

PS: I believe the content taught in Course 3 is also similar to Andrew's recent NIPS tutorial: Nuts and Bolts of Building Applications using Deep Learning. (https://nips.cc/Conferences/2016/Schedule?showEvent=6203)

There were many video editing errors and grammatical errors in the writing in this course. I felt an increased cognitive load and more frequent distractions because of this. I would be happy to volunteer some time to help with this if you contact me.

The pacing of this course felt strange; week 1 was very short and the material felt less important that week 2.

I would have really enjoyed a programming assignment with transfer learning. Since we already built a cat identifier, making a monkey identifier, or flower identifier would have been lovely and fairly easy, I think.

(09/2019)

Overall the courses in the specialization are great and provide great introduction to these topics, as well as practical experience. Many topics are explained clearly, with valuable field practitioners insight, and you are given quizzes and code-exercises that help deepen the understanding of how to implement the concepts in the videos. I would recommend to take them after the initial Andrew Ng ML course by Stanford, unless you have prior background in this topic.

There are a few shortbacks:

1 - the video editing is poor and sloppy. Its not too bad, but it’s sometimes can be a bit annoying.

2 - most of the exercises are too easy, and are almost copy-paste. I need to go over them and create variations of them in-order to strengthen my practical skills. Some exercises are quite challenging though (especially in course 4 and 5), and I need to go over them just to really nail them down, as things scale up quickly. Course 3 has no exercises as its more theoretical. Some exercises have bugs - so make sure to look at the discussion board for tips (the final exercise has a huge bug that was super annoying).

3 - there are no summary readings - you have to (re)watch the videos in order to check something, which is annoying. This is partially solved because the exercises themselves usually hold a lot of (textual) summary, with equations.

4 - the 3rd course was a bit less interesting in my opinion, but I did learn some stuff from it. So in the end it’s worth it. Not sure I would have taken it as a stand-alone course, though.

5 - Slide graphics and Andrew handwriting could be improved.

6 - the online Coursera Jupyter notebook environment was a bit slow, and sometimes get stuck.

Again overall - highly recommended

I found the course rather confusing than helpful. One of the key issues with video-only courses is lack of interaction of the user with the material. In previous Andrew's ML courses, this issue was cunningly tackled with "wake-up" multiple choice mini-quizzes. Such techniques would help the course a lot.

The questions in the exam were poorly phrased and full of typos; some had numerical issues (percentage of errors in the dev set did not sum up). Some of the answers seemed to contradict with the material as I remembered it from the course: f.e., the question on whether to get more foggy images to improve the model performance should have been answered with "augmentation is fine as long as it looks fine to the human eye". This contradicts to Andrew's remarks in the course video "Addressing data mismatch" video -> Artificial data synthesis. Are you sure we would not introduce a bias by adding artificial fog to frontal camera images?

Pros:

The course itself avoids many professional terms, so that most people can understand.

Introduced very valuable experience on how to think about the direction of effort and how to avoid wasting a lot of time.

Cons:

The problem with avoiding mentioning the program code itself is that although a lot of concepts are mentioned, there is still no sense of actual operation. A little simple example would help. It could be in the optional reading materials.

Use too many "so" for transition. Very confusing. Please use accurate transitions, or just pause it a while.

Often jump from what was talking about to something else suddenly. This is very confusing and hard to keep focusing and understanding. Maybe this is also because of the "so" problem.

Awesome course as always. The course teaches real world practical aspects of how to get started and navigate in the real world projects. The guidelines are actual learnings from years of experience.