Neural Networks and Deep Learning

Well planned, lets you develop intuitions about neural networks , also has optional video series on heroes of deep learning which is quite cool.

The Deep Learning specialization, which it is part of, is quite comprehensive too!

I really enjoyed it and highly recommend it for anyone interested on ML, Deep Learning and AI!

You probably won't learn everything there is in the real of Deep Learning just by taking this course (I think that would be impossible) but you will get a rock solid (trusted) foundation for the important parts to expand that knowledge and keep up with latest progress on your own afterwards.

It's a great course to understand basics of deep learning with a detailed walkthrough of gradient descent algorithm, forward propagation, backward propagation, cost and activation functions with logistic regression as a starting point.

very good matrial and also very good mentor Good introduction to deep learning.

Anybody who is seeking to get a knowledge of Deep Learning and/or Neural Networks then you just can;t skip this course.

very good and useful understandable lectures and hw Very good introduction in Neural Networks and Deep Learning.

However, I feel that not enough emphasis was given on how the derivatives were obtained, which seems like an important part of deep learning.

A great explanation of deep learning.

The must course for any deep learning beginner.

After learning, we can well master what is told in the course, and can learn further in deep learning.

Looking for more materials to study maths behind the forward propagation and backward propagation.

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.

I was intimidated by backward propagation before taking this course.

Especially, I have learned how and why the Backward propagation is working, which I could not understand before.

There is one point in the backward propagation in deep part of the implementation where dA of the previous layer is indexed by the current layer.

Nice, easy to understand implementation way of teaching makes learner grasp Backward Propagation algorithm in depth.

In my opinion, this course is pretty good.The content of this course be designed as a simple methods to learn and the homework was designed by your team ,it is very useful for me to understand this course and the progress of forward and backward propagation of deep learning .Thank you Andrew and others .

And then it moves to deep learning with neural networks with the techniques of forward and backward propagation.

very helpful to understand the forward/backward propagation in practice.

I believe students would learn more from a format where, for example, lecture notes were available in PDFs (e.g., containing forward and backward propagation formulae and, optionally, their derivations) and Python assignments were open-ended and less "step-by-step".

I loved the mathematical explanation of backward propagation, if gave a really nice intuition and pointed me in the right direction to pursue a more formal explanation.

The programming exercises are tiny toy examples, but help in illustrating the main principles of architecting the forward and backward propagation for NN training, and the vectorization techniques were useful and practical.

a great course Excellent intro to deep learning if you have a basic background in machine learning, especially in Logistic Regression.

Andrew Ng step by step explains from very basics of the subject (logistic regression) to very complex deep networks consisting of several hidden layers.

Also one open point: I still haven't understood why we use the log-loss function for logistic regression (and in all our examples) and not the squared-dirrerence... Just very clear and amazing!Cleare explaination, very good material and a funny teacher :) A must-have MOOC for neural networks apprentices So fucing brilliant!

Very helpful to start at logistic regression and move into deep learning to help paint a more holistic picture.

很好的课程，英语足够好的话，值得购买。 One of the best introductory courses, taking the learner from simple logistic regression, through computational graphs and teaching the core of neural networks viz.

But I would add that before doing this course, it is better to do some courses or tutorials on Linear Algebra, Logistic Regression and Gradient Descent so that you have a good foundation to understand this course.

And it also helps better understand machine learning concepts like logistic regression.

For me, the most fascinating fact was finding out that logistic regression can be viewed and understood as a very simple neural network.

inspiring covers the basic of "logistic regression", "2-layer NNs", "deep L layer networks", assignments in particularly very interactive.

The build-up form logistic regression to a deep network was executed very well, and allowed me to attain a good initial understanding of ANN's.

Starting from logistic regression, Andrew builds upon the materials and masterfully introduces the more sophisticated concepts one after another.

Here we spend 4 weeks to deep dive into it, comparing logistic regression, shallow nn (2 layers), and deep nn (many layers).

I also get a stimulus to recap calculus and linear algebra.

It helps if you are at least a little familiar with calculus and linear algebra, but Ng still does a great job of giving you the background info you need for these topics on the fly.

Highly recommended for anyone would like to understand NN and Deep Learning - no previous experience required although some basic calculus and linear algebra knowledge will become handy.

Andrew always wants to make you feel like "don't despair if you don't understand the math" but I do believe one needs a solid grounding on calculus and linear algebra to make the best of this.

After all, when talking about sigmoid / tanh / relu, explaining all the calculus and linear algebra behind it would be that complex and involved.

You do not need a background in Machine learning if you basic undergraduate knowledge in Calculus and Linear Algebra.

If you've never saw anything about Calculus and Linear Algebra, I think you will have much more trouble understanding the backpropagation.

Not sure how other people would fare, but I felt like in order to have a deep understanding of what was actually going on, I needed to go study the calculus and linear algebra behind the material (which I had done previously).

Thank you to all the assistants and TA's who put in so much time to this course!Now, for anyone who is debating taking this, having a calculus and linear algebra background will definitely help you in this course for the theory, but it's not a necessity at all.

ESPECIALLY for anyone who have been exposed to calculus and linear algebra, although he seems to describe everything needed.

It also helps one to brush up the calculus and linear algebra knowledge very much.

It blows my mind that using some calculus and linear algebra we can recognize photos with decent accuracy.

I've never programmed in python before and it was illuminating to see why its so popular in data science given the power of its math libraries.

Easy to follow for anyone who has some familiarity with data science concepts.

I will strongly recommend this course to everyone who want to spark in data science.

Now I set the vision to look into the data science and modeling the analysis.

This course is very valuable for the Data Science professionals.

This course is must watch for people who are or intend to use deep learning to solve data science problem to understand the logic of deep learnings.

However, I really need a class and especially homework to teach me how to derive back-propagation algorithm, which is important for job interviews of machine learning/data mining/data science in China Maybe the programming assignments should be a little more difficult, but overall I learned a lot.

In this respect, I think there are few better at getting to the core of teaching: simple is better.I appreciate that the course is designed to widen the reach of deep learning, but for those perhaps either more mathematically inclined or just extra curious, I highly recommend Lazy Programmer's Data Science: Deep Learning in Python class on Udemy for only $10.

These kind of courses have made me going really deep into Data Science and I'm quire sure this specialization will help.

this is a great course which explains in-depth the workings of a neural network , its highly useful for anyone looking to start a career in data science & AI Super helpful introduction to neural networks and nicely implemented theory in practice.

Highly recommended for Data Science enthusiasts.

A theory with applied programming, a Nice and well-planned course for those who have basic math, Data science or AI knowledge.

Great foundation to deep dive into the intricacies of artificial intelligence.

:) Superb...I am grateful to Prof.Andrew NG It was an excellent course that is extremely insightful for even people like us who do not have a good academic background of Neural Networks/Artificial Intelligence but are otherwise enthusiastic about this field.

I got introduced to the field of Artificial intelligence using this course and it was one of the best thing happened to me since long time.

I had previously took Andrew Ng's Stanford Intro to Artificial Intelligence Coursera course and this was a great followup.

It was perfect :D one of the best online course available.it got me launched correctly It was a wonderful course to get started with Artificial Intelligence and Machine learning.Those concepts of forward ,backward propogation, relu and sigmoid function was really new and helpful to get insight of what happens behind the scenes of machine learning algorithms many concepts were new and typical but Sir Andrew did a great effort and explained them in a way that everyone can understand it.

Thanks a lot Prof. Andrew for awakening the knowledge in Artificial Intelligence.

It's a very important course if you want in to the Artificial Intelligence, the professor Andrew is a very specific in his explications and great teacher.

Great course for beginners Very interesting and knowledgeable course for beginners in Neural Network and artificial Intelligence.

First of all I would like to say that this is a wonderful course for the beginners who often wonder how to start and enter into the world of artificial intelligence.

this course finally helped me to establish myself in Artificial intelligence field and it was a correct pathway for me .

Andrew Ng is one of the world's leading practitioners of Deep Learning and Artificial Intelligence, and he's also an excellent instructor when it comes to explaining concepts in a simple, lucid manner.

Therefore I believe this Deep Learning course can help me to possess the basic ability to work in the field of artificial intelligence and deep learning.

Fortunately, the chain rule (otherwise known as forward/back propagation) is hardly one of the most difficult derivations in machine learning.

For me, calculus is not a problem, but for others who are not familiar with calculus and its chain rule, his detailed explanation would help.

期待着证书的颁发 Very good explanations especially for the derivative intuition and chain rules explanations which makes my life easy in understanding back propagation.

I would've liked some more details about the derivations themselves, although, thankfully I'm still able to do the maths, but I think they might help some people who are more curious by nature but aren't particularly familiar with chain rules or derivation in general.

Good course, learn about the derivatives chain rule and how that relates to back propagation algorithm.If you want to understand the formulas of the derivatives, you must do the calculus yourself.Some ready-to-use Python implementations of an L-layer NN.

I would suggest some more backpropagation/chain rule emphasize early in the course (as in CS231n) as this is so fundamental for understanding the latter parts of the course.

As a math nerd i was hoping to see a little more in the chain rule derivations but that definitely isn't critical to the successful completion of the course.

(chain rule etc.

Derivation chain rule, for example, is hardly explained and is critical for really understanding what's going on computational graphs.Second, programming exercises are too simple, and most of them are about copy-paste and not about real understanding.

Went through Derivates, Chain rule , matric multiplication.

Also, it would have been nice a derivation of the backprop algorithm for the general case (maybe as an optional video, where you show how to use the chain rule for several variables).By the way.

Buen curso para poder conocer las bases de las Redes Neuronales.

Buen curso, se aprenden las nociones básicas de la inteligencia artificial y las redes neuronales.

Buen curso para adquirir un entendimiento de los conceptos fundamentales de las Redes Neuronales y Deep Learning.

Muy buen curso, te da la base para entender como funcionan las redes neuronales.

Gran curso, unas expliaciones muy buenas sobre las bases de las redes neuronales, con una buena base matemática.

fantástica introducción al universo de las redes neuronales desde un punto de vista tanto divulgativo como académico.

It is a great course with the perfect balance between practical and theory :) it'a a very great and organized course.

Perfect balance between hands-on and theoretical approaches.

Excellent course, perfect balance between theory and practice.

Perfect balance between simplicity, detail, and applicability!

It provides in-depth knowledge of mechanisms behind neural networks such as gradient decent, forward and backward propagation with perfect balance between intuition and math proofs.

The exercises are the perfect balance between student involvement and fun.

Having taken Geoffrey Hinton's Neural Networks for Machine Learning, I still consider the programming assignments to be very challenging but there are plenty of materials that helped me getting through it.

I find this course as a potential companion course for Geoffrey Hinton's Neural network lectures which provides you deep insights about various network types, and you can hear many stories about the early days of NNs from a cognitive scientist's point of view, but lacks of hands-on coding and practical exercises.

The masterclass with Geoffrey Hinton...priceless!!!

Great going guys The optional material with Geoffrey Hinton was a treat.

It also makes a few questionable decisions such as putting a 40 minute interview of Geoffrey Hinton at the end of the first week, most of which you will not understand unless you've seen neural networks before and have familiarity with his work.

A little too much hand holding, but overall quite useful!

Much recommend.One small thing I think could have helped a bit is the practical examples do a little bit too much hand holding.

深入浅出，很赞！ very well taught Good overview, but too much hand holding in terms of programming assignments.

I feel theres too much spoon feeding in the assignments.

A little less spoon feeding would have made the assignments more interesting for ex refer lecture X for the back ward propagation formulae etc The course servers as a perfect intro into Deep Learning.

Not being a computer scientist, and with my coding skills a bit rusty, I was grateful for what some call "Python notebook spoon feeding".

The assignments were good, though I feel a little bit less spoon feeding on the ipython notebooks would have offered better learning experience.

This course is more of spoon feeding, I liked the introduction to neural network in "Introduction to Machine learning" course better.

Excellent A little bit too much spoon feeding, but good for the basics.

spoon feeding for beginners, forget about these stuff quickly Really interesting, even if basic, new knowledge.

Assignments - a bit too much of spoon feeding.