This is a practical course, extremely helpful for those who have met so many troubles in realworld projects. It is quite helpful for startups, where we can implement those ideas immediately. On the other hand, the transfer learning and end-to-end learning paradigms might be very useful but challeging in big companies and sectors.

This is one amazing course because it exposes you to a 'real' ML/DL problem. As a newbie I learned a lot and hope that in future I will once again do it as a ML research/development Engineering Manager. This is something very practical and now while doing big projects I will consider the learning of this course. Thanks Andrew Ng.

Exceptional, one of a kind strategic course for ML practitioners. The amount of wisdom and knowledge shared in this concise course would definitely save any budding ML engineers from the common pitfalls that many teams may still face. It also sets the foundation stone for cultivating prospective machine learning project leaders.

the best course course so far in terms of (error analysis, how to deal with training/ dev/ test sets and what the symmetry of distribution means, how to split data set in the best way, how to be able to use an algorithm again in another deep learning project, how it's important to correct the incorrectly labeled data set, etc )

Consider this a course on best practices. I found fundamental advises on how to best carry a ML project from scratch, regarding the first model you should choose, how to perform on different scenarios, how to choose systematically your train/dev/test set and so on. The project simulator is a must, I wish they put more of those.

Really really good course. I never knew about the intricacies of error analysis that is done in ML/DL projects. This was a very insightful course. Would see the lectures again if I need to (which I will). Nevertheless, amazing course. The content is explained in a step by step and appropriate fashion for even a newbie like me.

The teaching in this course is so invaluable for interpreting the results. Now, I believe I can understand my models' accuracy based on professors teaching. The professor teaching contains unique knowledge and experience, where you can't reach via the internet, library or asking your university professors. Thank you, Prof. Ng.

I think this course was very valuable in teaching insights about how to think about and formulate ML/DL problems. The case study quizzes were really good and made you think. I hope coursera expands on these case study quizzes for future version of this course as well as introduce them into other courses of this specialization.

This course is a summary of Andrew's experience. I've yet listened this nuts and bolts from Andrew speech(you can find it on youtube) but there are some precious advice that are so much valuable. I'll recommend this course to everyone who want to start a carer in DL. Big thanks to Andrew, the Deeplearning.ai team and Coursera.

Plenty of wisdom shared by Dr. Ng here, presented in a very digestible and actionable fashion; can't wait to apply to approaches suggested to my own projects. These kinds of courses are golden, can't find such practical knowledge in ordinary textbooks. Thank you for the course, can't wait to continue with the specialization!

This course taught me recipes about conducting a machine learning project. I'm now more confident about being a machine learning project lead. The assignments are interesting because they are case studies of real situations, where decisions need to be taken in order to iterate and converge to a better machine learning model.

This course really help me to understand exactly how to make decision to distribute the data sets, what to do with the new data set, how to examine the error, how to use previous model as a transfer model for other classification, what is multi-tasking and many more

Thank you for your support and sharing of your knowledge

:)

This course was most useful as Andrew explains practical engineering challenges and valuable tips to overcome them!

As a technology architect, I am more interested in predictable, guaranteed results and can guide my my ML engineering team to make the right choices in given real-world uncertainties and engineering challenges.

This course provides us an overview of the errors we have to encounter while solving a machine learning problem and shows us a clear direction of overcoming those. Though the contents are not mathematical but these information will help us to deal with machine learning projects in efficient way. I really liked this course.

The course is very practical and also leads you to learn the real challenge you will encounter while working on machine learning project. While it's easy to follow as the previous courses, you need to think more strategically. I would recommend bringing an idea or a project you are planning and apply what you learned here.

The course really streamlines and puts forth a structured approach to go for delivering a machine learning solution to a problem. It helps to complete my project in 2-3 months instead of a year that sometimes some of my colleagues take. They need to look at this course. Also the interview with Ruslan was very informative.

Really liked this course, especially the case studies, where the task is clear and possible scenarios are explained. Have to response in the most promising way using the knowledge obtained during the previous 2 courses. Really appreciate this experience. Only wish is to have more case studies in the other courses as well.

Great course. The pragmatic insights were invaluable. I think addressing problems such as missing input data and data preparation would help. I also think a programming assignment that explores these ideas would help. You could take the sign language number exercise from week 2 and explore some of the ideas this week.

I can confidently say that this course has content which is only unique to this course. To my knowledge no other course has topics like Avoidable bias, Bayes optimal error, Error analysis and emphasis on train, dev & test set data distribution mismatch. This course is definitely a must for any Deep learning practitioner.

Looking at the title of this course I predicted that it will be regarding to teach me how to organise the source code files of ML project and more specifically how to build a ML project and components of deep learning project but it was all about DEBUGGING ml project so for me this was in off beat course from its title.

Extremely useful, practical techniques for deep learning projects. I feel much more able to construct my own neural networks, diagnose and solve issues with them after following this course. Professor Ng is a gifted teacher. His style is careful, methodical and never less than very well prepared and deeply enlightening.

Its really nice to get the valuable insight of managing an AI project, this course not only thought us about deep learning, but also how to manage them efficient and take smart decision. I like the concept of Transfer learning as it can same a lot of efforts and time to build an system for complex. Thank you very much.

I am very lucky to be able to learn from Andrew the DOs and DON'Ts of how to develop a successful practical deep neural network for real applications. It would take a machine learning developer many years of working experience to acquire any one of the topics that Andrew articulated in this course. Thank you so much!!!

I completed the entire specialization and having nothing but good things to say. Highly recommend it! Lectures are engaging, and Andrew does a fantastic job explaining some very complex topics. Programming assignments are challenging in a good way. You’ll really feel like you’ve learned a lot by the time you’re done.

The course is great. It tackles a lot of problems regarding strategic decision making and at the same time important concepts such as human-level error, avoidable bias, transfer learning, end-to-end deep learning and others are being taught. The questions/exercises really test the core concepts that are being taught!