Slightly disappointed with the assignments to be honest, most of them are too easy to solve, and moreover can be just copypasted from the hints.

Great theory which lacks some demanding practice tasks.

Seems lacking some mathematical details like how to calculate Backpropagation this time

Excellent course, although it would have been good to talk more about backward propagation, after finishing this course this is the only point that is left unclear in my mind.

Just as in the other courses in the specialization, this provides solid conceptual overview, but: 1) too little conceptual practice (the quizzes are short and inanely easy) 2) too confusing labs (for someone with plenty of JS, particularly web-related, but no prior python/data-science experience, I struggled to understand how the data structures were being manipulated and to intuit what the helper functions were asking me to do, and plenty of numpy/python errors that I had to interrogate for too long, distracting from the core concept. On the latter point, I laud their attempt to scaffold with hints, but this needs revision).

I was really hoping more detailes on xgboost and unfortunately the course level was very elementary

I finished this course with 3 weeks left in my monthly subscription, only to find out the 3rd course wouldn't be available until the day my subscription ended. I essentially paid for a specialization that didn't exist at the time of my purchase.

Great course! and according to me, the ML roadmap that best matches the one I thought to approach the ML topic based on all my experiences. So I recommend this course of Andrew to everyone.

The content is excellent but some more emphasis must be given on the discussion of the codes in the Jupyter Notebooks otherwise it'll become less appealing to the once who don't have a good grasp over Python.

Enjoyed thoroughly the course. The mathematics concepts we’re well explained through exercises that help us visualize the concept behind each equation. The exercises were well thought out to help the student bridge theory to practical coding. People not familiar with mathematical coding will start to understand the pattern behind and be self-sufficient. Thank you for building this wonderful course.

This was one of the best courses I have ever experienced. There was a subtle beauty in the course's planning and Dr Andrew's teaching. The effort Andrew sir made in his teaching was quite evident, and there was a remarkable balance in the difficulty of the course - no matter whether you are a beginner or experienced with Machine learning, you will enjoy this course!

A very beginner friendly course that has great explaination on the topic. The programming assignments are really simple and I hope to see harder assignments in course 3 which would allow us to put everything we learn into practice

The course was fantastic! I really enjoyed every part, every video, every quiz, every optional lab, every assignment of the course. It was a pretty memorable ride to have come this far.

Good introductionary course to advanced learning algorithms.

Main point of feedback: the course did not address in anyway how to actually desigining neural networks, i.e choosing layer size, number of layers, etc. I thought this was a bit odd, since this is a rather fundamental part of NNs. Now that I've finished the course, I'm still not able to build my own NNs from scratch, since I don't know how to choose my layers. I hope further courses will address this, but I think this should've been addressed in this course already.

Nicely presented, however the assignement and the material was too basic.

I would have liked a more rigurous introduction to the subject, that being said I feel I did get something out of this course.

There are way too many errors on the video's subtitles - not grammatical mistakes, but words completely out of context or even incorrect formulas.

The course expands the subject of machine learning in an appropriate and interesting way. Unfortunately, the quality of the downloadable material is insufficient, as the course describes multiple graphs and equations with numerous superscripts and subscripts, but unfortunately it is not possible to download the material in the slides, so as the complexity increases it becomes very difficult and stressful to take notes. This results in passive listening to the video, which certainly has a negative effect on real learning. Finally, each exercise contains the solution and it is therefore possible to pass the course passively without having learnt anything.

One of the best explanations of the backend of most advanced ML algorithms. The instructor is the best of the best. Learning new and Advanced Algorithms surely helps you to understand the newer technology and help your hunt down a good job in your future. Thanks to Coursera, Stanford, Deep Learning.AI and Andrew Ng for this very special course and for all of your hard work in making this course go smooth and easy to understand.

It is a great course. Only one issue is that the lab assignments don't require much effort (are kind of feel happy problems) which is not good for learning. I didn't bother to check the details of assignments as that was not necessary for writing the code and pass. The assignments should test all the stages of making a ML project, right from data collection, feature engineering, training, validation, etc. Thanks!

Amazing, the explaination of the neural netowrks part is amazing wtih very good examples.

The Bias and Variance and the error analyis parts are very good.

I loved giving an idea about more advanced implementations like transfer learning and ensemble learning.

Really good introduction to machine learning notions and algorithms such as neural network and decision tree, with many informatives details about how to improve them and examples of implementation.

Another fantastic course by Andrew Ng! He covers neural networks, decision trees, random forest, and XGBoost models really well. I like that he shares his intuition behind every concept he explains.

excellent course, for me it's the best online course you can see anywhere

Interesting content! I feel like I'm learning a lot, and I appreciate Dr. Ng's friendly, straightforward explanations. I wish there were more hands-on programming practice; the practice labs etc. often feel like most of the code is written for me and I'm just filling in a line or two. Most of the learning feels like a "sit and get" format rather than a hands-on experience, which makes me nervous about how much knowledge I'll retain and be able to transfer. I also think it's odd that, after introducing vectorization in the first course, the labs consistently seem to push us to write implementations using loops. That concern aside, I am thoroughly enjoying these courses and find them extremely helpful and approachable.

Good exposure to ML concepts but the labs were a little too easy. I think if they ever redesigned the course they should give the student the option of coding the algorithms from scratch.

I think it gets a little messy at the end but aside from that it is a very great course that I would highly speak of an recommend to anyone considering a similar path