I would love some pointers to additional references for each video. Also, the instructor keeps saying that the math behind backprop is hard. What about an optional video with that? Otherwise, awesome!
Really, really good course. Especially the tips of avoiding possible bugs due to shapes. Also impressed by the heroes' stories. Genuinely inspired and thoughtfully educated by Professor Ng. Thank you!
By Mohammad G H•
Very basic level
By David B•
This course is really quite bad. I'm not sure why the rating is so high. Probably because they are only prompting people who completed the course to rate it.
The main problem with the course is that It spends the majority of its time describing a byzantine set of notation while avoiding actually helping you understand how to apply the concepts you're learning. So you learn that a^[l](i) is the activation vector for layer "l" and example "i" but then you get to the python portion and, big surprise, none of that information is even slightly useful.
Even worse, the course hasn't chosen its audience. If you're good at math you'll be annoyed about the math explanations. If you're good at programming you'll be annoyed by the programming explanations. Rather than isolate that material in a way that lets people skip parts which they already understand, you get a really basic explanation of everything all globbed together.
Anyway, I'll still try to hack through this thing to finish it, I'm just letting you know that if you're underwhelmed, you're not alone.
By Richard R•
Meh. I don't know why we are spending so much time in Week 2 talking about the math and how to not use FOR loops in week two when he STILL hasn't given any kind of overview about why we do this math, how we're going to use it to identify cats in pictures. Instead, we're just yakking on about math math math math math with NO context whatsoever. If I wanted a math class, I would have taken a deep-in-the-weeds math class. I expected a higher level of instruction for this higher level of abstraction but instead it seems that he just wants to talk about math and how to use vectors in NumPy. Zzzzzzzz.
By Andrew H•
Not enough explanation or support to complete the very vaguely worded assignments in anything like the specified timescales.
I respect the source of this course but as a teaching resource it is really very poor.
By Bedrich P•
Course teaches bad programming practices, such as naming variables dZ and b. Also it is little outdated - neural networks are not written in numpy anymore.
By Ali A•
Terrible integration with Jupyter Python framework, end up losing 3 hours of work! Nobody responds from the courser team !
By ABIR E•
The course is more than excellent, you will implement all the Artificial Neural Networks Algorithms (step by step), and you will learn all the maths behind these algorithms. The Assignments (especially those of the last module are challenging!). I already obtained the Professional Certificate "TensorFlow Developer": now I understand the behind the scenes of many packages of TensorFlow... really: the course is terrific!
By Tim B•
The course does not have the same quality as the “Machine Learning” course Andrew Ng made with Stanford.
The biggest issue are the programming exercises, that do not require the learner to think at all. Most tasks in them are on the level of “copy and paste this piece of code”, “retrieve a value from a python dictionary” or “use a mathematical formula displayed directly above”. I appreciate the effort to make the course more inclusive to people with a weaker background in Computer Science. It would however make the course much more worthwhile to have challenging exercises with optional hints, instead of giving the solution away in each task description.
“Neural Networks and Deep Learning” hardly teaches anything, that wasn’t already covered my “Machine learning”. The major differences is that it uses Python instead of Octave and arranges features as rows instead of columns. In my eyes, the learners time is better spent, skipping the first course of the Deep Learning specialization entirely and taking the Machine Learning Course instead. To the creators / maintainers of the course I would advise creating a summary, that covers the most fundamental differences between the two courses (different notation, numpy fundamentals) and make a suggestion where someone who has taken Machine Learning should join the Deep Learning specialization.
While the audio quality has improved, the video editing is poor. There are multiple occasions where misspoken content, that was clearly meant to be edited out, remained part of the video. Many videos are preceded by a “Clarification” reading task that corrects some mistake in the video. How hard is it to get an intern to fix this in post?
By Anne R•
The programming assignments provided a good framework in order to practice coding the main functions in a neural network. This was helpful to understand the matrix operations underlying the forward and backward processing in a general L layer network. Without a previous background in linear algebra and in neural networks however this course would be challenging and maybe very frustrating due to the limited debug information available.
The course videos need to be a lot more focused on the details being conveyed. The verbal and visual discussion and explanation provided is in my opinion not effective. The slides are cluttered and contain many errors, the verbal portion is like a casual conversation that repeats quite a bit, and the script provided for those that get tired of the repetition contains many transcription errors. I would recommend that someone be paid to correct the scripts to help those that prefer this way of working through the course material.
By Ofer B•
Very abstract, and the examples are not as concrete as they could be. I'd use better visuals to ensure that the concepts in each video are understood 100% visually.
By Miriam G•
Really just mathematical background knowledge. Nothing you would ever need, since there is keras. No own thinking during assignments neccessary, either.
By Aratz S•
Easy course if you have coursed the ML course before. I would like to see more explanations in detail. Still some bugs in the assignments... why???
By Thomas M•
Course starts with a lot of math without any context what all those computations and parameters are used for or what they have to do with N
By Loren Y•
The assignments are not good. Too easy and too much handholding. Also lots of technical issues.
By Tobias G•
Few Detail. Mathematics missing.
By Gaetano P•
The course is well structured and the explanation is linear and mostly clear, but:
1- in 2020 I expect that in doing such a course are going to be applied relatively modern teaching standards, like for example avoiding handwritten text. What is the purpose of writing on the screen if you can use animations to more clearly connect concepts during your lessons?
2- I don't expect that errors to be just rectified before the video. Reupload the video? Errors like that during long formulas and explanations are just going to kill the learning. It is pointless to write before the video that in the future video you will make an error. Just correct it ON the video.
3- If you can't explain in-depth calculus, just to di with the help of someone else. You cannot exclude calculus.
4- The only thing i've learned in this course is vectorization (thank you). The rest is just copy the formula given during the explanation (handwritten on the screen.....) and paste during the exam. I didn't learn how to apply a neural network because during the "exams" it was built already. I expected assignments to make me build an create every piece of the network, instead it was all already done and all i had to do was repeat what Andrew says in the video. This is NOT learning. You need an assignment per video for that kind of thing, you can't just go forward and write some formulas on the screen pretending you have "explained it" because nothing seems explained to me. Why should i use those methods or formulas instead of others? Nothing is explained.
This course is really good but assignment given to solve is not understandable.
By Kenneth T•
Great course, definitely taught me the basics of Neural Networks and Deep Learning as it's supposed to. Assignments are quite engaging when you try to thoroughly solve them. Even with minimal mathematics, the course will handhold you the whole way. Definitely a great course for anyone with minimal programming to get into. For me, the most challenging part was understanding how Python syntax worked with numpy. If you are taking this course I recommend taking your time with implementing the projects, they can definitely give you an understanding behind the logic of neural networks by following the code. The instructor is quite nice and warm, sometimes a bit dry, but nonetheless, he seems very warm; wanting to teach the next generation of individuals to do ML/AI. The course does have a few downsides such as how buggy the iPython notebook can be. This is the programming environment you will be using. An the video quality isn't always the best with the audio, but overall the content was presented in a great way and prepared in a manner in which you learn one step at a time.
By Sandip G•
The content was very good and intellectually curated, and no complaints about a teacher of such high quality "Andrew Ng". Actually, I took the "Machine Learning" Course long before on Coursera from the same instructor, as I took this course now, which highly helped me to finish this in less than a week, although I never got time to complete the former course. Advice to any new students on this course would be to have a basic understanding of Machine Learning, which includes linear regression, vectorization et.al. , (or simply, "ML" course on Coursera).
One small amendment on this course could be to reshuffle the contents a little, from different weeks as I found the content which was in Week 4, to have high importance to be taught earlier in this course (for eg, getting matrix dimension right ), and there were others sub-topics in week 3 as well. I don't remember all of them, as I took 4 weeks worth of information, in just a single week :)
Very excellently taught, and contents, as well as assignments, were of topmost quality.
By Kenny C•
One might dislike that the derivation of formulas is not talked about in this course, but I think it's the right decision for this course. I took the Coursera Math for Machine Learning Specialization before taking this course, and the derivation for the formulas took at least 4 weeks of background material about linear algebra and multivariable calculus. Thus, this course aims to give you a conceptual understanding of neural networks that will allow you to implement it on your own. While some might argue that the programming assignments are too easy, or that too many hints are given, I think they're necessary for guiding you in the correct direction during the assignments. If you take the time to read the prewritten code, you will be able to get the understanding you get from writing it fully from scratch and possibly taking hours to debug and to read NumPy documentation. Overall, a very solid course for those who want to build a neural network on their own.
By Irfan A M•
Learning from Prof. Andrew Ng (Stanford University, founder of Coursera, an eminent researcher in the field of Machine, Deep Learning & AI & founder of so many lead companies in AI) indeed Blessing.
Such a composed course you get a chance to learn the underlying concepts of AI, Machine & Deep Learning, and implement real-world problems to get intuition and exposure. The design of course content and relevant assignments develop your concepts deeper and intuitive.
One of the prominent features of this course was listening to Heroes of Machine, Deep Learning & AI; Prof. Geoffrey Hinton, Prof. Pieter Abbeel & Prof. Ian Goodfellow really give you motivation and intuition about latest happenings and future directions these fields.
By Michael C•
Excellent course. Surpasses Andrew Ng's original Machine Learning course in conceptual depth and ease of implementation. The lecture videos, quizzes, and programming assignments are all targeted towards someone who knows nothing about deep learning or machine learning, yet manages to elaborate on surprisingly advanced topics which you would not expect to make an appearance in an introductory course. It strikes a superb balance between simplicity and depth that is rare even in in-person university courses, and much rarer still in MOOCs. I will be taking all the rest of the courses in the Deep Learning Specialization. Well done.
By Hong X•
I've learned to build the basic binary classification model from conventional logistic regression to a shallow model (with one hidden layer) up to any layers of ANN. One of the most rewarding point for me is that I start using python (other than Matlab with which I have stuck for years until recently most cutting-edge open-source codes are found delivered in Python!). Although there is still a long way to go , I found well warmed up by those delicately designed step-by-step programming exercises in Jupyter notebook. Therefore, I do appreciate the course materials contributed by the lecturer as well as the exercises-designers!
By Chi W C•
Wonderful class. I started out not knowing anything about neural network or deep learning. I was able to follow the class lectures to get a sense of what was going on. The assignments were clearly structured and well organized, and serves as excellent examples in how to build this type of applications (by small building blocks and test each of the block carefully).
At the end, I was able to build my first neural network implementation in recognizing a cat!!
(However, I have uploaded 3 non-cat images, but NN failed by predicting these were cats. On the contrary, logical regression correctly predict the 3 images as non-cat).
By Carl G•
Andrew Ng is a thorough teacher and shows how online platform can be as engaging as taking a live class. His pace and style of writing slides is perfect for keeping pace taking notes by hand (my preferred way for efficient learning). He takes time to explain in depth how NN's work, and even more important his experience how to use them. Homework is a bit simple, but also appreciate to not be mired in coding details. Nice to be able to focus on how NN's works. Best part is that each piece of code can be fully tested against known output before used further. Illustrates nicely good practice once doing real coding project.