Thanks Andrew. I really enjoy this course. Although there are plenty of knowledgeable lecturers in Coursera, a few of them know the teaching skills like Andrew. I specially took your course since I had taken Andrew's ML course in 2013 as well, and I knew how patient he is and how well he can teach to anybody with different level of knowledge. Assignments were challenging but clear. The checkpoint helps a lot and make sure learners if they made a mistake, they would not lost. Merry Christmas and Happy New Year <3

Really good course. Andrew is clear, and provides a great introduction to structured deep learning. I feel that some extra videos showing the full calculus behind back propagation would be helpful for those who want it. However these can be found elsewhere on the internet if you look around. Assignments are pretty good, with a few things I would nitpick - however as long as your methods return the correct outputs, you can rewrite the internals as you desire. Heroes of deep learning extra videos were interesting.

I've taken other machine learning courses before and I was somewhat familiar with neural networks before taking the course. The error backpropagation technique was something I couldn't get my head around intuitively. But, after finishing the graded exercises in this course, such ideas are well cemented in my mind. Andrew does an excellent job of controlling the mathematical jargon from overwhelming beginners. I would definitely recommend this course to anyone who wants a hands-on experience with neural networks.

In my honest opinion, I found this course to be extremely well organized. The concepts put forward are done so in a manner that does not overwhelm most of the students but it also provides just enough challenges to those of us that are interested to further explore the algorithms and the mathematics behind them, if need be.

Andrew Ng is by far my favorite educator of all time and I feel truly blessed to live in a time that allows me the opportunity to learn from the wisdom and know-how of such brilliant minds.

Definitely a satisfying course for beginner. If you don't have any experience on neural network and deep learning, this is an excellent place to get started. Videos are great with full explanation on topic and Professor Ng keeps telling "don't be scared by math or python notations". I spend about 6 hours for first coding assignment and learned a lot on details to get thing right. After that, I just feel much better on following assignments. Cannot wait to start course 2. Thanks everyone to provide this course!

Fantastic course. Andrew has always been instructive and can explain complicated things in a simple way.

The assignment codes are very well structured and with the skeleton outlined it's not that difficult to finish the homeworks. But building from scratch is another level of challenge and I'll try that after the course.

Also the back propagation mathematical induction is not covered here. The math part is also good to learn when I have time, but I guess I'm more interested in the application of Deep Learning.

I am a fan of all courses delivered by Andrew Ng. This one gives a thorough understanding of shallow and deep Neural Networks. After the course, you can be sure to have a sound understanding of how a model is built from scratch. The assignments are also organized in a way to reduce your effort on redundant tasks like creating the structure of functions. The focus is only on having you write the code that tests your understanding. I really enjoyed this course and I am looking forward to the next one. Thanks :)

Following the lectures and completiong quizzes and programming assignments puts the student through rigorous math, the math in the course is overwhelming at a college math/engineering level - but as the student continues throughout the course, the redundancy of theory does get absorbed and by the end of the course the student develops a solid understanding of the course material. The programming assignments implement full scale deep layer neural networks with practical applications to illustrate the concepts.

This was an amazing with a lot of new and interesting things to learn. I am really glad that I decided to take it. Its approach toward neural networks is quite easily understandable and allows oneself to use those concepts as he wishes. The programming assignments are a really big help as well. You can learn all the math but without the programming skills, there is hardly any point in doing deep learning. A special thanks to Prof. Andrew Ng. I was already a fan, but this course was just amazing. Thanks a lot.

Excellent course if you want to understand deep learning at a high level without delving too much into the math. This is exactly what I wanted because I am an engineering manager who has to make decisions about ML/AI. I have a math background and can see how some of the math works, but what I wanted was a more visceral sense or intuition about deep learning, and this course is perfect for that. The math actually gets in the way, and Prof. Ng very cleverly kept the focus on the intuition rather than the math.

Great course!!! The best online course I have ever taken! I enjoyed almost every day I participated in that course, really an educational treasure! It is so comprehensive and detailed at the same time. Due to the good presentation of the topics it was really understandable. The only thing I would wish for future participants: please make it easier to get the complete Jupyter notebook environments from the Coursera platform once completed. I spent a lot of time here - even after consuming the related blogs.

Great introduction to neural networks. STRONGLY recommend pushing through to week 3 and 4 material. Weeks 1 and 2 may feel slow at first if you're familiar with the material but Andrew Ng goes into a lot of good detail in the equations in weeks 3 and 4 and if you try to understand the code you're writing in the code blocks you'll get a good intuition for the forward propagation -> cost functions -> back propagation -> update params (the learning part) neural net loop. Excited to jump into the next course!

The intro course has been a phenomenal experience learning. The concepts were clearly explained along with derivations. I thank Coursera, Andrew Ng and all others who were involved in this for taking this massive step in teaching deep learning and AI. I would be happy to take more practical oriented courses under this banner especially computer vision, NLP, AI in specific. Another suggestion from me would be to include lessons on building neural networks from libraries like tensorflow, pytorch, keras etc.

Very helpful course. Great, well prepared assignments! Even without python knowledge I was able to code essential parts of algorithms. Practical assignments were really good reward at the and of each week and a motivation for me to keep going. You will not be forced to learn python in parallel but occasionally I have to read library reference guide to debug. I also installed python locally to test syntax and get more in, but it is not necessary, provided python jupyter notebooks is also usable for this.

Thank you for the easy-to-follow content. The explanation about back propagation in details is great. The Python code is elegant and should be a good starting point for learners to make more progress in expanding it.

Some time assignment submission gave errors even there is no problem with networking issue. This definitely need to be improved, or learners need to resubmit many times.

If you need translation of the course to Vietnamese language, let me know. I will do it for free, for my Vietnamese students.

A great introduction to neural networks! The videos and assignments were helpful, and the repetition helps things sink in. I would've preferred more mathematical rigor and a little less hand-holding in the assignments, but I understand that this course is meant to appeal to a wider audience and it does a good job of being approachable. I particularly enjoyed the weekly "Heroes of Deep Learning" videos, and tips and pros/cons of studying machine learning in industry vs. academia. I'd recommend this course.

It is a relly nice course, well explained as Andrew Ng. has always done. Because it is still a new course, there are few erratas of course, but those are being already corrected. I suggest a lot to take the Machine Learning course by Stanford University here on Coursera first, unless you already know about Neural networks, since sometimes there are things that you should know. These kind of courses have made me going really deep into Data Science and I'm quire sure this specialization will help. Thanks !

Thanks a lot Dr. Andrew NG. According to me, this course ranks very high in terms of course content, delivery and practical assignments in Python. Specially, the assignments are designed in such a way that almost all the concepts are revisited and the conceptual understanding is re-intensified during the assignments. The assignments also help us to understand how a neural network can be implemented in practice in a systematic way by breaking it into subcomponents, which is the most enriching experience.

Excellent class, wonderful instructor and extensive practice problems. The theoretical explanations on deep networks are very thorough with the math behind it. Unlike other deep learning courses that take shortcuts with using pre-made keras or pytorch libraries, this course went through the math behind the functions and then went on to build them with python from scratch. The exercises are also well prepared with clear notes and test functions to make sure the codes work as intended. Highly recommended!

I hope instructor could fix the notation in back prop. I think this should be easy, because he just need to stick a red color comments beside in the video.

One big misleading is by back prop:

Because the notation for back propagation algorithm presented in the lectures treats dA and dZ differently from dW and db(I ignored layer l index in my notation). Namely, 'dA' and 'dZ' are always computing the derivatives

dL/dA and dL/dZ

respectively, but 'dW' and 'db' are computing the derivatives

dJ/dW and dJ/db.

Again an awesome course ,hats off to NG for this brilliant series of courses.

One thing which i liked so much was the interview session with Ian ,Peter etc.Came to know about further research and development going in Field of ML & DL.

i liked the way Ng has put up the lucid explanation of vectorized implementation and how to do random initialization.

And the ending was super with DNN for image classification.

Its a good experienced learning so far with Prof Andrew.

Thanks & looking forward to next course .

I took Andrew's Machine Learning course but was never able to complete the course. This time I have completed this course and hope to complete the remaining 4 as well.

Andrew has been very successful in developing the intuition for the neural networks and once it becomes intuitive it's all imagination.

I loved all the interviews with "Heroes of deep learning". To be honest, I never knew about any one of them prior to those interviews. It is great to know the best people in the industry.

Thank You Everyone

I have taken other courses in Deep Learning but this was by far the best one since Professor Andrew Ng explains everything from the ground up and also provides most useful intuitions whenever needed so we can apply them to other problems. Other highlight was the fact the programming exercises wanted us to implement the network from scratch without use of any ML libraries and it greatly improves our understanding of the concepts and was not the case with other courses that I had taken. Highly recommend.

Certainly challenging for this guy but I stuck with it and I am glad I did. I have a background in engineering and little python experience so the learning curve wasn't terrible but there were certainly some challenges. I've been in the oil and gas industry for nearly a decade and looking to change careers (hopefully somethin in ML). I'm glad I took this course because while challenging, it solidified my interest to seriously pursue a new career path. Worth the cost and I'm looking forward to Course 2.

a basic course, given the depth of mathematics it discusses. One good thing in the course is the frequency of the practical assignments, however, I feel the course needs one small project where each student writes the whole program on his own to get used to the whole process, rather than just implementing the functions. One thing I believe needs to be added, is to offer hints as an Optional thing, so that some people feel challenged (as well as grasping the idea in a deeper way hehe) during the course.