Andrew also has a good teaching skill. After completing this course, I made a very simple data set myself, and then calculated the whole forward, backward propagation procedure with pen and paper. Though it was very simple data set(L=2, m=3, n0=2, n1=3, n2=1, g1, g2=sigmoid ...) and I iterated only two times, it took really long time far more than I've expected. But It helped me a lot to have a feeling that I really get a sense on this!

Simply the best designed introductory course on Deep Learning I've ever seen online! Professor Ng is really charismatic and presents the material in a simple and concise way. Also, the homeworks are very well organized and definitely helps the student to absorb the main concepts without struggle with python programming hurdles. I really enjoyed the course and I'm looking forward to the next course in the Deep Learning Specialization! :)

I was afraid of learning deep learning because I thought that I do not have sufficient Calculus background to learn deep learning. However, this course taught me good enough Calculus, and also made me realize that I do not need to have significant knowledge of math to learn deep learning. The learning was fun and exciting due to the way Professor Andrew derived the equations, and the way the course, along with exercises, was organized.

Course materials clearly explained as usual by Andrew Ng. I actually seem to just go through as if it is a review because I took his ML course (online) for Stanford but the most excellent part I like here is the Jupyter notebook and the flexibility of the deadlines which was not as easily movable as it was back in 2017. Finally a realistic feature that busy people who got real life emergencies can avail of without dropping the course!

Another brilliant by Andrew N G. It introduces you to the world of deep learning in a step-by-step. A very carefully artciulated course which presents the overwhelming concepts of Deep Learning in a simplistic manner. Key terms & methodologies used in the world of DL have been introduced & explained. This course truly desmystifies the concept of "Deep Learning". A key higlight is the interviews with the gurus of the Deep Learning world

The explanations are great! Simple enough to understand it without leaving out important details. I had already taken the Machine Learning specialization from university of Washington here at Cousera, so I don't know if the explanations would be equally easy to understand if this were the first time hearing about logistic regression and gradient descent... if you are not completely new to machine learning, this course is great for you.

Great at giving you a different perspective at how an L-layer NN works as a sequence of independent blocks that you can add to have a deeper NN. Also, it was very enlightening to finally know what I am doing, in each step, when using the backpropagation algorithm, instead of just computing the deltas without knowing what they are.

I would only ask for a brief introduction to python , specifically, what is a list, dictionary and a tuple.

Andrew and his crew is again doing a pretty amazing job! I found it incredible how he is capable of transforming complex topics in such a simple and straight forward easy to follow steps. I have bachelor in math and master in statistics but I still learn something new in each and every video. He is always showing a new perspective of everything. Simple, elegant and full of content. Thanks for putting the effort on such a great program!

A perfect course to learn deep learning from the scratch. First, learned about implementing Logistic regression to classify cat vs non-cat images. After that used Neural Networks, learned several activation functions, cost function, updating parameters and so many other. Used NN to train a model to classify between can/non-cat images. This course has built my fundamentals of Neural Network. Thank you Andrew Ng for this amazing course.

This course is recommended for all the deep learning beginners. The structure of this course is very good and gives a step wise building blocks for building a simple neural network. This combines a overall good theory and there need to be an improvement in the programming assignment as they are not very tough.

I really gained a lot of knowledge from this course and it has helped me build neural networks with good theory and confidence.

I think this course is the best course I've ever had.

Professor Andrew Ng is the top instructor in the world and said all the topics in the easiest way.

I highly recommend this course for each person who wants to start machine learning and deep learning.

Thank you so much dear Coursera for letting me have this course.

I hope in the future I will be able for your platform to make courses and give it to all people around the world for free.

I really like how this course was handled. The concepts were presented clearly and in "manageable chunks", and I understood them easily. The programming exercises were also a good complement to the lectures. Also, even though I am a beginner in Python, I was able to complete the exercises since the programming exercises test one's understanding of the concepts than mastery of the programming language. Thank you for a wonderful course!

This is definitely a five star course. I've read a lot about deep learning (books, blogs etc.) but this is the first time I build a neural network myself. I learned a lot and I do not agree with the 1-2 star reviews. It's true that you're doing some stuff again and again, but that's what learning is all about. Concerning the programming assignment, then they are not super difficult BUT at least now I know how to build NN from scratch.

Excellent and clear introduction to subject. Prof. Ng encourages with great patience and systematically lay out the principles, step by step in lectures. The assignments are very well set out, illustrative of NN and Deep Learning principles with practical examples. They are challenging too. A basic level of Python programming ability is helpful to complete them.

Thank you Coursera and Prof Ng for another excellent learning experience.

This Course was very well structured and a planned course.The programming exercises gave a guided tour of the application of the concepts.I deem it as a perfect starter material for learning "deep learning".The Video transcript requires some working though.The sound quality can be improved.I had to focus on the transcript the entire time to listen carefully to the lecturer instead of watching what he was actually doing on the slides.

The fundamentals of a neural network is wonderfully explained by the professor in this course. I found this approachable even with little prior knowledge in Machine Learning. The calculus required for the course is light, though I found the discussion forums rich with threads about the optional calculus parts. The hands on assignment helped me get a good feel for implementing basic neural networks from scratch in a modularized manner

Best course to get you started in Deep Learning. I think learning everything from scratch is extremely important, even when you can go and start coding straight away, because having intuition about what is happening in the background will surely make you more comfortable about your code. Also, making your code from scratch helps you innovate and make changes to the slightest of details that could improve your model to a great extent.

Programming assignments was really good and well explained walkthrough to make sure student understand the linear algebra. Adding and intermediate level of two layer N and then L layer networks helped a lot to see the scale and really understand the concept. I think Professor Andrew is made for teaching, he really knew the art of explaining most complex problems to student. Thanks a lot for the Entire team for putting this together.

Overall, it is a very good introductory course in Deep Learning. Professor Andrew Ng is definitely an expert in the field and make the explanations very clear for everyone.Personally, I would like to have more difficult programming assignments (maybe build the whole functions by ourselves) or even better to examine if there is a way to start doing some projects in this kind of online courses (similar to the ones of Stanford's class).

I am very pleased with the structure of the course. I just wish the programming assignments had more 'coding' for the students. Even though i got a solid understanding of the concepts through the programming assignments, I wish they had been more difficult, at least without the hint giving away the solutions or the no of lines of code. Apart from this, I loved the course. It has given me a step from which I could explore DL further.

Clear explanation of the deep learning concept. The programming excises were tailor-made to reduce peripherals codes writing, which allowed us to focus on the core material coding. Good job on the team! This course cleared two concepts I was puzzled before. 1. When should use sigmoidand other activation functions on Forward/ backward propagation. 2. Backward propagation related to gradient. Thank you Dr. Ng put together this course.

Frankly, at first I think this course is less difficult than UMich's EECS598, but when I fully completed this course, I am amazed that Andrew Ng. can illustrate things so clearly. Despite its simplicity, it indeed teaches me everything I need, and provide me with deep insights into this field by interviewing other famous people in DL field. I would like to express my gratitude to Andrew Ng. and continue to finish all the 5 courses.

After finishing Machine Learning courses, the most suitable course to continue our study is the specification course series of Deep Learning. As the first course of the series, Neural Networks and Deep Learning contains not only detailed explanation of neural networks but also practical training written by Python. The course only spends four weeks to finish and it is easy to pass. I will continue to the second course. Let's do it!

It's the best way of explanation I have ever seen. Also, the assignments are very helpful, not too hard, and not too easy. Just lead us to the right code to help us learn. Interactive questions that come after the videos are beneficial, they make us attentive. To the learners, in order to get the best out of the course, I suggest to go through it twice or more and put everything you learn in a notebook and solve equations yourself.

Excellent introduction to the topic. I particularly liked that the topic is introduced from "first principles", i.e., including the calculation of derivatives (for gradient descent) and the implementation of back- and forward propagation functions. I would, however, recommend to others to re-implement all training exercises/networks in your own python programs to further practice and test them for yourself (and find your own data)