This course is very informative and useful for anyone who have no machine learning background. The case study approach helped a lot in understanding the core of every concept before deploying.

The only drawback was that there was no tutorial on how to install the software which was so tricky and take me ages to install it.

The course is pretty decent but what i really didn't like was it outdated use of software and pretty less efficient mentors.I suggest, to please moderate this course, this course is indeed a good one but need to be supplied with new references and less dependent on particular libraries.Otherwise the course was great.

I think the course met my expectations – it's super high-level and does not at all go over the underlying algorithms involved. I give it 4 stars because I have this feeling like this specialization is an underhanded way to sell the Dato GraphLab Create product. There's a bit of a conflict of interest going on here.

Wass a great introductory course. Definitely recommend for starters. The course was well constructed and presented. The only problem I faced was from the software side. I was having a hard time installing and importing packages, those are not covered in this course. I like the case study approach as an introduction.

The course has everything you need to get an overview of machine learning. It's perfect to understand the purposes and techniques used. However, I'm a bit concerned with practical tasks, since they heavily imply on GraphLab create, and this is a serious disadvantage, since one will barely use it in future.

This course is great if you just started getting into the field of machine learning. (Great if you have no or limited programming background)

Pace is a little bit slow and Programming assignments does not captures algorithms discussed in lecture.( Although it mentions, it never let you implement yourself.)

Very good introduction to machine learning, quickly enables the student to perform regression, classifications, etc. but it would be nice if the course went into a little more detail, the quizzes are very superficial. It would also be beneficial to explore examples of applications across different fields.

I enjoyed this class. It does provide a good over-view of the different machine learning algorithms and their practical applications. My only qualm was that the programming assignments seemed somewhat irrelevant as the underlying structure of the different machine learning algorithms had not been taught.

Nicely covered basic ideas about different areas in ML . Hans-on sessions gave a very good idea to solve ML problems practically. Theory explanations where good.

One suggestion i had is about tool used it would have been cool if course was with scikit learn and pandas, since many companies use the same.

I really enjoyed the overall materials and especially loved the way they split the course into two sections:Theory and Programming.

However, i wish they were using more standard packages such as pandas, or skit-learn instead of graphlab. It was a good class to taste what is going on in the world of ML.

Great introduction to machine learning concepts with nice assignments. It seems there needs to be some cleanup performed so that the lectures and content match up a bit better. Overall a useful and approachable course to motivate the need for additional study in the rest of the "specialization."

I really enjoyed this hands-on course with a lot of practice. The difficult part was the week 1 when we had to set up the virtual environment, and pass the first quiz. I believe that if there is more support at the beginning, more people would have stayed on to finish the course.

Thank you!

Assignment instructions are not very clear and often not formatted properly - multiple questions are "glued" in a single block. It would be good to spend more time on GraphLab API at the beginig of the course. iNotebook did not work on coursera.org, I had to install a local version of Python.

The course has a really interesting approach, and I have enjoyed it, but as Turicreate (previously Graphlab) has not become a mainstream tool (on top of the difficulty to use it with Windows), I would rather have also some exposure to other tools, as Scikit-learn and Tensorflow, for example.

Case study approach is really helpful but we need to understand the formula behind those deep learning scores which i think has not shown in this course. Second, if you can provide more videos for coding then it will really helpful to do the "Programming Assignments" which i think is tough.

Well designed and executed in the main. As videos are recorded once but viewed thousands of times it would have been nice if they had taken more time to write a script and look more professional - also mistakes in the videos that don't match the current material are few, but very annoying.

Good Course for someone wanting an overview of techniques. You would not be building something very cool after this. Course assignments are not very challenging, some good questions must be included.

Some mathematical part should be there. Expecting it in other courses in specialization.

I found this to be an excellent course. Both the instructors are excited about the work they are doing and do a good job of teaching the materials. I don't have a statistics background and college calculus is more than a little fuzzy. I was still able to understand most of the course.

Pros:

1. Practical and hands on approach

2. How multiple problems like prediction, sentiment analysis, text retrieval etc can be mapped to a common ML model

Cons:

1. Uses DATO toolkit which is a licensed tool.

2. Maths behind any used technique is not discussed, which I think is an issue.

Very nice for those of people who want to learn the concepts of machine learning. However, libraries that used (which are not pandas and scikit-learn) was not satisfying for me. Googling about those libraries was totally difficult while there are lots of documents using pandas and scipy.

A great high level overview and introduction to machine learning. The topics broadly cover different machine learning algorithms and their approaches. The only downside is that the proprietary library they use to teach the concepts is incredibly expensive for an individual to license.

It easy to start and content is less mathemetical and ready to use knowledge. And tool are helpful to start playing with. May be later go into details . It's perfect for software engineer not having knowledge of ML but wants to start. One can play with data after this course I think.

This was a great specialisation taught by great professors, but I'm very frustrated that the final segment on deep learning was never finished! Coursera needs to fill this gap since it is hugely important topic and isn't covered adequately in any other Coursera material right now.

The course is designed very well, nice way of teaching taking practical examples , it is just that working with pandas wasn't easy,it took a lot of time even to find syntax and graphlab wasn't working tried searching on various forums. Please add ski-learn and pandas syntax too.

Pretty good bird view course for entering the world of machine learning. I liked how the course was designed and its contents. It provides us the machine learning ways of thinking and doing, and get us ready to move further and deeper in the path of machine learning. Recommended.