Very good and well structured. Easy to follow and it is great for self-learning (although you never know whether it's you or the neural network). The assignmets are very motivating and easy-to-follow. After taking this course, I DO think I understand what is happening inside a deep neural network and I feel I could code one from scratch. I willing to keep learning-deep :))

Very well presented. Starting with a familliar model to illustrate the mechanics of deep learning really helped some of the more technical aspects sink in once we started scaling the size and layers. I think there could be less guidance in the assignments to encourage notes and demonstrate mastery of the material, but overall I got a lot more out of this than I expected.

Excellent and thorough introduction to shallow and deep neural networks! The lecture videos are in-depth, easy to follow, and provide enough mathematics direction for those with little calculus experience and for those with a good deal of it. A beautiful connection among linear algebra, calculus, and computer science. I am moving on to the second course. I am very excited!

I've always enjoyed Andrew Ng's course starting from his Machine learning course broadcasted in 2011. This course was particularly a beginner level AI course. It would've been better if the coding exercises were a little bit harder. All the steps were given in advance which made it really easy to complete the entire exercise within 30 minutes eventhough I am new to Python.

!!!! The best class on neural networks!!! Strong recommend!!!

Thank you for all teachers and students who make this class for us online! It's really a great class and programming exercises are great! I can really learn something and begin to know how to build my model! Thank you, Andrew Ng! and thanks for coursera that let me have chance to learn this class. (Financial Aid)

The course Neural Networks and Deep Learning was a great experience for me. Professor Ng and the mentors are very knowledgeful and professional. The second week was quite hard but still manageable concerning theory. The programming exercises are really cool and at the end the student has a running framework to implement a deep neural network. Definitely five stars of five.

Excellent course & many thanks Prof. Andrew Ng! The parts that I like the most is the clear & easy-to-understand explanations for gradient descent, forward & backward prop. Compared to other online resources where the explanations went lost in thick jungle of maths, Prof. Andrew Ng managed to keep the maths at the right amount but not losing touch on the necessary details.

Very interesting course, It's easier to start with this one than the bigger Machine Learning course from Andrew (obviously, it's covering less material than this one). Being fluent in Python and Numpy will certainly make you feel more at ease but there is a lot of material to be able to cope with this difficulty. Can't wait to complete next courses from the specialization.

Excellent course materials make the neural networks not mysterious any more. From easy examples to more complex ones, from equations to coding exercise, the student can gradually learn the inner working of neural networks. This is a much better approach than starting with those big daunting back propagation equations. Andrew Ng's teaching was as good as his Stanford days.

This course is awesome. You are going to build your own neural networks, with multiple hidden layers, on your own in Python. The things that are best about this course is that it will not just teach one how to implement the neural network but also the proper way to day that, with vectorization. The assignments are pretty easy but very helpful in understanding the concepts.

This course has been very carefully thought out to make it as approachable as possible. Andrew Ng is an amazing teacher, taking difficult material and making it seem clear, intuitive and reasonable. The integration of Jupyter Notebooks for the programming exercises is brilliant: I spend zero time fussing with infrastructure and 100% of my time learning the material. Bravo.

As an intro course (following on from Andrew's ML course) this does a great job of demystifying the maths and intuitions around forward propogation (weights, biases and activations) through the neural network layers, and then the calculations around the gradients for these during the backward propogation phase. Vectorisation is a clear winner in enabling the processing.

I think it would be better to add a few more video lectures talking about the assignment after finishing them. Some parts of the assignments, especially when it comes to the dimension of variables in the code, become particularly confusing. But overall this is a great course that I will definitely recommand to any person who wants to start their journey of deep learning.

This is a great course for me who have no idea about DL. NN Concepts are explained until tiny details. Math seems not that difficult but also explained well especially derivatives part. For now, I don't understand all codes meaning in programming assignment, although this is good to get insight about how NN works. A good foundation for you who wants to step further in DL.

this is an excellent course for programmers who want to delve into the deep learning world. Professor Andrew explained everything intuitively. Even one who is not good at mathematics also can do this course. even though Andrew explained it very easily you have to do some work and read, watch and rewatch the videos so that you could get a good understanding of the concept.

I loved the learning experience. Complex topics have been covered in a very step by step manner that is easy to digest. Being familiar with calculus helped me understand math and derivations of the formulae used. I would suggest adding additional (optional) videos that dwell a bit more in the math behind backpropagation. Otherwise, I am highly satisfied with the course.

You'll get the most of this course if you took Andrew Ng's Machine Learning course, at least until the weeks that covered neural networks. Thank you to the course instructors, TAs and organizers. Andrew Ng can teach machine learning to a brick, he delivers the content with clarity and simplicity so much so you're learning complex concepts without realizing it.

Had a blast!

Great introductory course of neural networks. Some more mathematical details would have been nice, but that is outside the scope. Programming assignments were easy (assuming prior knowledge of a programming language). I feel I know enough to start reading some literature on neural networks, and understand the basics of whats implemented in packages like caffee and theano.

El contenido teórico es muy profundo, pero introduce los conceptos de forma muy repentina, sin incluir referencias previas clave que ayudarían a una mejor interpretación. En mi opinión, se detiene demasiado en conceptos de cálculo de una o más variables, que deberían conocerse de antemano. Sin embargo, las prácticas son muy completas y refuerzan bien la teoría del curso.

i am a math-physicist with lots of experience in C and other programming languages, but no previous contact with neural nets, i really enjoyed these lectures, after which i could program my own small NN from scratch on my PC. These lectures are nice and the exercises at the same time easy and very enlightening. I strongly recommend this course whatever your background.

I had a very elementary knowledge of Python from a college course I took 6 years ago. Luckily, my linear algebra skills are much stronger and carried me through most of the course. It's VERY well done! After just four weeks I feel like I have enough to start my own DL algorithms. I can't wait to see where I am at the end of this specialization. 5 stars, highly recommend!

I liked the course. I even learned a bit of Python ;-) I'm not a machine learning expert, so cannot really judge the content. I have a suggestion though: make the programming assignment grader more verbose, i.e., provide the input and expected output when the student's code is deemed incorrect. The input and expected output can be generated randomly at each submission.

It was an excellent experience learning this course. The teaching methodology, structure of the course and short crisp chapters make the learning very easy and certainly it can be learnt easily by students from non CS and non mathematics background also. The programming assignments and quizzes are such that these provide excellent practical understanding of the subject.

Excellent introduction course in Neural Networks concepts and Deep Learning delivered by Andrew NG. Highly recomended for anyone interetsed !

Would be beneficial for students to have the slides accompaning the videos, especially the ones oultining the core concepts of the model architecture and components, vectorization forward/backward propagation, and gradient descent.

This is one of the best courses if you want to start learning deep learning from scratch, all the maths has been explained simply and intuition on how neural nets work has been conveyed briefly. Understanding deep learning from basics allows one to practise it on programming frameworks easily hence i am very thankfull to coursera and Andrew NG for putting up this course