Another excellent course by professor Andrew Ng, short, on point and clear videos.

There are no programming assignments in this course, the skills learned on how to debug / lead a machine learning project are being examined through 2 graded quizzes that simulate real world projects.

It is worth mentioning that some interviews with influential people on the field of DL are included and make the student fall in love with DL even more. Excited to see what's next in this specialization.

This was a good course, overall. It covered a lot of the decisions you need to make, when configuring and working to improve your neural network models.

There are not actual flight simulators. That is just how some of the learning exercises are described.

This course made me think about a lot of things--for example, is it better to simulate noise in "clean" data or to try to filter noise out of noisy data. Obviously, this course is just a stepping-off point for your own explorations.

This course felt a bit out of sequence in that it left behind the more "hands on" notebook coding for a higher level "How to manage an AI team/project". This made sense when I realised it used to be the last of a three-course specialization. Aside from how it fits into the flow of the specialization (which then moves on to get technical again with CNNs and RNNs), it's jam packed full of incredibly sound advice that even experienced team leads would probably benefit from reviewing.

In this module professor Andrew NG teaches several strategies based on his vast experience to help you deal with real world machine learning projects. Most of the information is of great value and it is difficult to find organized like that in another website. I have really enjoyed the two case studies proposed and they are very interesting to help you review the concepts studied. Finally, professor Andrew NG explains the content clearly and it is a pleasure to watch his videos.

This is another awesome course teached by the best instuctor (prof.) in the net for ML and DL technologies. Knowing how to divide the dataset in the appropriate sub sets and doing the right error analyisis, is the main goal every developer or scientist in this field tries to achieve. This course teaches all this and additional concepts like transfer and multi-task learning which are essential techniques to improve productivity. I would give six stars if there were any.

I feel humbled as I ended this course, realising that years and years of knowledge that Prof. Andrew and others have gathered they've just let out to public, accessible to everyone. It is such a great act of kindness. I am really thankful to you folks. This was a great course to learn the insights of an experienced ML / DL guy. It would help a lot when I'll actually be working on a real life project. I hope I would be able to spread the light of knowledge even further.

This may be an optional course in the deep learning specialization, but I beg to differ. If you plan to do actually start a project in machine learning, it is imperative you take this course. You could finish your project ten times faster with a fraction of the work. All in all, every one of these courses in this specialization have been beautifully organized and taught very well.

Thank you so much for offering this course along with the others in the specialization!

This course is quite amazing! Currently, I use lots of frameworks like Tensorflow, Pytorch and some NN architectures which are already open-source and available, I could know what's the general problem with my project and also how to fix it, how to find the most promising directions for my team. It's like now, you know how you can improve you model precisely and know how to start a new project ( What sould we do ?). This course guides me a lot! Thank coursera!

The course is really great! It offers an in-depth understanding of the practical aspects and applications where deep learning can be applied. Most importantly the content in this course will help you iterate faster with your machine learning problem by doing error analysis on it. This course tells you exactly what can be done in which situation to improve the performance by analyzing the data and other statistical aspects of the data as well as the algorithm.

Error analysis and Learning the methodology to handle the errors. Besides the traditional systematic way of performance analysis like train-dev-test and cross-validation, Andrew focused on data mismatch and train-dev data. These two are the most important things that are described very well. Like other courses, Andrew was very good to describe the real-life practice. In this course, two simulation quiz really helps a lot to deeply understand the application.

This course mainly explores the ideas related to good practices and guidelines while executing different types of learning, and does justice to them. Overall, it is a nice course. It's not too heavy of a course compared to the first two parts of this specialization and I could reasonably complete it in a week. Moreover, it does not have coding exercises so keep in mind that the goal of this part is to introduce certain ideas related to design of projects.

This is a VERY valuable course, with tons of practical advice on how to understand problems all machine learning / neural network developers experience and how to tackle them. I have never seen such high quality practical advice in any textbook or in any other course before, and I believe that even those who are not taking the full five course deep learning specialization should seriously considered this course. Another truly excellent Andrew Ng course!

This course introduces some general principles for developing a deep learning project. It points out the difference of setting of train/dev/test sets between deep learning and traditional machine learning. That's a practical advice. And it's notable to include human performance and regard it as Bayesian bound, almost the best we would expect an algorithm to achieve. That saves you from spending unnecessary time to make a subtle improvement. Learnt a lot!

In this course I learned about ways to approach some of the real world challenges that I have already faced on some of my own projects. For example, what actions should you consider when you find a significant number of labeling errors in the dev/test sets that affects your ROC. I also was motivated by the last module on end to end training and the interview with Ruslan Salakhutdinov to pursue an end to end training idea that I have been thinking about.

A very valuable and must-have knowledge for any Machine Learning and Deep Learning practitioner is covered in this course. The "how-to" of analyzing and handling of different situations based on error rates is very practical and is based on real time experience, and might not be covered in other courses or subjects. Also, great coverage on Transfer Learning and Multi-task learning. I have learnt a lot and feel more confident now. Thank you Andrew Ng.

Great Course! Many students will choose to skip this course since they think there are less knowledges than other course in the 5-course specialization. But I have to tell you: this is the best course in the specialization, because you can learn a lot knowledges especially skills and experiences in practice from this course that you can't learn from other books, courses or universities. BTW, I'm not telling that the other 4 courses are not important.

Andrew Ng is brilliant! I have never seen such a great tutor in my life. He bring extremely useful concepts and explains them so easily in a way the concepts stay in your mind.

Like the backprop algorithms he talks, he has learned so much from his old course and he has made great improvements to focus on New people. He sure has a good deep network up his brain that has gone through lot of iterations (without overfitting) with beautiful set of features.

One of the rare courses to learn about structuring our own deep learning projects. I found the material I learnt in this course very useful in my carrier as well. All the subject matter are well structured and the flow of the module is very easy to follow and understand. Together with the case studies, it was very enjoyable and very easy to test the applicability of the knowledge gained. I highly recommend this course for any deep learning enthusiast.

It is very good probe to practice. I would very appreciate to take this course before I have started in machine learning. It would help me to realize some mistakes I have maid before. On the other hand, for people, who have some experience with machine learning, some chapters are being over-explained, as the topics are quite clear to those people. Overall: I would recommend this course to everyone, who wants to start with his/her own NN training.

No coding at all. But this is one of the best course on AI, because it does not talk about coding or anything, but most importantly, the one thing that is not taught by many others. Experience of Andrew Ng trials and errors in approaching ML projects. How to create structure, how to observe what results to see. In short this course is like 'how to save time in doing AI projects and make optimal use of it, avoid trial error which can cost months.'

Despite of its relative simplicity (from a technical point of view), it is probably one of the most practical courses I have taken in Coursera. Even though it only mentions deep learning, the overall methodology can be applied to any machine learning work. It is important to get familiar with the heart of the models, but it is probably even more important how to work on an end-to-end machine learning project. In summary: Highly recommendable!

Hello learners, and respected teachers ! As I go through this course, I was not clear about how to use test set. train set and dev set, was not able to rectify about how to break data set for so closely prediction. As we have teacher like Andrew ng, we won't miss anything about structuring machine learning project.

This course is really helpful for me and will recommend to all learns who want have command on ML & Deep Learning.

Unlike most of the Deep Learning knowledge which can be found in literature and other MOOCs, this course provides you with insights that can only be acquired trough (often painful) trial and error. Here you learn how to approach Deep Learning projects, how to avoid most common mistakes, and how to quickly identify errors in your model.

Do yourselves a favor, and finish this course before taking on your very first DL project.

This course is my favorite so far. It has really given me a way to systematically and strategically set up a machine learning experiment and iterate in a way that make sense. For me the toughest part of ML projects has always been figuring out where and how to start. Now that I have some solid guidelines to follow, I don't feel as anxious about jumping into a new problem and it turning into a wild goose chase. Thanks a lot!

Such an amazing course. This course should be done by every Deep Learning researcher or enthusiast. If the previous course taught how to debug Neural Networks then this course teaches how to perform advanced debugging on the dataset. The contents of this course are invaluable and not taught in most institutions and they are not known to the wider community that actually follow the guidelines and techniques presented here.