The theory provided by the course is really good and I will highly recommend it as an introduction to ML. The practical applications are not as good. They feel rushed and not particularly well explained. You usually get a recipe that you follow through using a tool that hides most of the details behind the theory. You just feel like you would not be able to do the assignments without the tool.

Loved the enthusiasm the instructor show while teaching! However I would suggest either to update the course video to consider Turicreate or change the assignment back to Graphlab, as it can be troublesome. Also installing graphlab on windows was a big challenge (at least for me), would really appreciate if there was one video explaining how to do that!! Overall content wise it was nice.

Excellent course! My only "complaint" really is that I wish the instructors would use other tools besides GraphLab. I realize that this does make it a lot easier, but I'm sure many of us work in environments where we cannot use GraphLab (for a variety of reasons) and it would be most helpful it they demonstrated how to accomplish the same thing solely using open source alternatives.

Love the professors and the curriculum. The video design is quite annoying though -- if I want to rewind to a particular place, oftentimes I would accidentally press the configuration buttons because they're so close to the timeline(?) (I'm not sure what it's called). Also, it would be great if the subtitles didn't have a background color, as it blocks out a lot of the video.

Great introduction to the various topics of the specialization. Introducing each of these topics in a generic way helps to get a sense of what they are about, and what to expect from the specialization. It might be a good idea to be explicit about the fact that this course really only has value added in conjunction with the other courses, and not really as a stand-alone module.

It was really good.

Pros:

Has really nice assignments.

Teaching is really good.

Cons:

Should've used and open source package. Graphlab is good, I accept, but I wasted like 4-5 hours trying to install it, because some or other errors or dependencies,. I mean some consideration should've been done about an easy smooth method for it, for a beginner like me, it was really frustrating.

The installation process of jupyter and ipython notebook can be more clearly explained because many people including me had a tough time figuring out the right way to install and get started with the assignments. Many people are stuck there only for 2-3 days and give up the course.

Otherwise I enjoyed the course with the different approach of teaching(the case study approach)

Very good course and inspiring case studies! A small fix needed is to change the order of the concluding remarks of each week and the Jupyter Notebook exercises. In the film you always refer back to the Jupyter Notebook exercises, but when I look at the film (following each week very linearly) I have not yet done the exercises. Thank you for a fun and interesting course!

I really enjoyed the intro, I would suggest that you take a bit of Python, and Brush up on your Calculus, because you will be needing it. It's possible to keep up conceptually, but I barley made it, and am taking some Python, and Calculus to make sure I'm not waisting time learning skills that are needed in the course, and can focus on the material being presented.

This is an awesome course for those who are curious of what machine learning is all about and want to get a broad overview. You don't really have to have programming background, just some basic math will suffice. However, if you look at some background and practical implementations, Andrew Ng's course is the one to go. This one is too shallow to anything practical.

The materials are very interesting. Great introductory course. There are a few problems that need workaround if you are using Amazon AWS to complete the quiz. Need some FAQ section for people who are using Amazon AWS. Wish there are some documents reference python/learn. Excited to take the deeper course but need to upgrade the computer instead of using AWS again.

Very good introductory course, especially to someone with little knowledge of ML. My only frustration is that I have some background in machine learning, so not very much new material- I would have skipped this and jumped straight into the the regression models course, but you need to complete them all to unlock the capstone (which looks really, really cool!)

This course is very practical for starters in Machine Learning. The ML methods it teaches are very fundamental, with clear and intuitive explanations about how and why they work. This course is recommanded to those who want a hands-on learning experience and who are not statistics experts that do not care much about the mathematical proofs of the algorithms.

Great overview of machine learning concepts. The content is structured in a helpful way to give you conceptual understanding of machine learning. Some of the quizzes and materials are a bit messy (most of the challenge is figuring out what parts are out of date or related to course material that previously existed). For the price, it's definitely worth it.

The course was really informative but is a little outdated as it uses Graphlab and SFrames which is available in the older versions of Anaconda and Python. It is also a very tedious task to study this course on windows. I had to install Linux on my system to study this course. I gave it four stars because it cleared the concept of ML through Case Studies.

In general, a nice "into" style course to show the capabilities of different ML solutions. However, trying to be so "cool", "awesome", and "exciting" slash back easily: the first thing I remember from this course is that annoying squeaking giraffe, and not the real content.... Nonetheless, If you need a practical introduction to ML, its worth the effort!

Although is a good for a start, and Graphlab framework is state-of-art software. I would like the same content developed using mainstream opensource frameworks like pandas, scipy, numpy, etc.

One of my interest in this courses was that they used python instead of R (which I don't like too much for it's inconsistencies). Python is faster and more reliable.

though it was a great course i was a little let down when module 5 and 6 were taken off the specialization series. A great course for beginners to understand machine learning as it introduced the aspects of it without getting too involved in math so that we could grasp the basics first. definitely will recommend to those wishing to dip their toes in MI.

Course is good starting point into machine learning specialization, it does not provide a good insights into the concepts as such but provides a good overview of different machine learning concepts and keeping the user interested in the course at the same time. I would recommend pursuing the course if you are starting into the machine learning domain.

I learned good material but it was very confusing getting started with graphlab create. Supposedly there is a switch to Turi Create, which I received an email that stated '..the content in this course has been updated for Turi Create', but none of the content is updated. I still got everything to work and do feel I got what I wanted out of the course.

Great course. Good for a broad overview. If you already know basic concepts like regression, classifiers, etc from a statistics class it might be a little slow. Also, the class is taught using graph lab which is not a free software. If you wanted to try it something else that is free, you could... But you will spend a LOT more time on the assignments.

it was great tho the week 6 quiz was quite difficult. a lot of comments in the discussion forum were helpful for me to complete the quiz but lots of feedback suggested improving the lessons to match the quiz or vice versa.

i think it would also help upon submitting the quiz to display the answers you chose , not jsut whether they were correct or not

Fantastic course! Great teachers and very nice to see real-world applications in action. Would have rated 5 stars if an open source library like scikit-learn were used. Students can still use sklearn, but all the examples are using GraphLab Create, a great library, but you need a very expensive (at least for my budget) license to use it commercially.

Before taking this course, I read users' reviews, I knew that this course is a bit out-dated and to my surprise, it mostly uses the proprietary library, graphlab, turicreate. But those obstacles didn't stop, I vowed that I'd learn the concepts but implement the exercise and other things using open-source packages, like Panas, Scikit-Learn and so on.

Very good introductory course on Machine Learning. Be prepared to dedicate extra time to explore the turicreate API. Overall well packaged quizzes and exercises. I found the explanation of math in some areas (for example recommender systems) somewhat lacking; however there are further courses in this specialization which might cover things in depth.