There should be a note at the beginning of the assignment in Week 4, that we may run out of memory with the auto-grader and what to do in advance to avoid that. My biggest time in Week4 was spent looking for and upload umpteen times (trial and error) to find a memory problem instead of upload to learn to calibrate parameters. Received 0.81 (which is rather ok) in the end but the distaste remains.

The course is a good introduction to ML. It covers lots of basic supervised ML techniques. The lecture slow pace is appropriate for presenting complex issues. It would have been beneficial to spend more time on the python case studies that are barely explained. Coursera platform issues with submitting and grading assignments should be highlighted in the assignments; not embedded in the forum.

Provide a quick and good overview of important, popular machine learning topics and their practical use with Python scikit-learn module. The material covers the important parameters to keep a watch on for performance and highlights the usual pitfalls and missteps. Very practical learning, makes one comfortable using ML tools and quickly apply for real problems like in the last assignment.

The course is designed perfectly and the pace is such that beginners in machine learning would enjoy. The course was well structured out and in a span of 4 weeks I think i learnt a lot. The only limitations i found were with the autograder not detecting files and other minor glitches like the videos not being marked completed even upon completion. But those can be fixed easily.

Just like other couses in this specialization, this course has great assignments which help alot.

As to instruction, totally different to previous courses, this instructor covered almost everything, probably too much for a four week course. I think I start to have some sense of machine learning however, I do need more study, probably Andrew Ng's course and refresh my maths.

Materi dan penjelasan yang diberikan selama kursus sangat bermanfaat dan mudah dipahami dengan baik, ditambah dengan pembawaan materi yang terstruktur mempermudah dalam memahami penggunaan dari setiap algoritma dan parameter yang ada. Kekurangan yang ada hanyalah di bagian penilaian assesment yang bermasalah diakibatkan dengan file yang tidak terbaca.

I learnt a lot in this course. The lectures, assessments, and reading material were all top-notch. The forums are immensely helpful. However, I'm giving 4-stars rather than 5 because I spent more time than was necessary in overcoming autograder issues. My suggestion is to improve the autograder so that assignments can be submitted more seamlessly.

I really got a lot out of this course. I started with a solid background in traditional data analysis (PhD in experimental physics), but knew nothing about ML. This was a great overview, providing a just the right trade off between depth and breadth--plus it was short, which is good. I can now go and do deeper dives into the material. Thank you!

The combination of assignments and lectures worked niceley for me. Good feedback on the discussion forums, too. Only thing which should be improved is the auto grader. The course introduces a lot of algorithms, but also gives you insight into how to evaluate their performance. In the final assignment it all comes together, which is always nice :-)

I think it gives a great overview on Machine Learning and Sklearn. Nonetheless i noticed it is less curated compared to the prevoius courses in this specialization (wrong filenames, unfunctioning links, old version of pandas respect the one used till now). Anyway it worthed and I'll give a look also at the optional unsupervised learning part

The teacher's voice is not motivating, it made me fall asleep all the time. But content is surely good. It's a perfect checkpoint after Andrew Ng's machine learning courses, by making experimental practices over theoric practices. Seriously, speaker needs to speak more alive! I don't want to hear deep breathe noises when watching a course :)

I Took this course blindly without knowing much about data visualization libraries. It took me a month or so to learn them first and then attempt this course further. The course study material is very decent but the assignments are pretty good and tricky. It is definitely a must-go-for course and I would surely recommend to my colleagues.

This one is very good and informative.

Although there is no explanations how to decide what type of preprocessing do on data set (to choose whether or not to do winsorization, convert categorical features to one-hot for linear models and to labeled for trees, etc) it still very helpful in understanding of PRACTICAL part of machine learning

This course was very interesting. Probably the longest course (duration wise) in this specialization. This course had to cover a lot of ground in 4 weeks time. Thoroughly enjoyed the assignments and it was challenging as well!. Gave 4 star because there are minor problems wrt. Autograder. But content wise there are no complains.

Lectures were a bit slow, I personally felt pace could be increased and more content could be covered in areas like boosting and all.The assignments gave me a hands-on approach in using sklearn library.I felt it was over-all a very good course and would definitely recommend it for others.

Thank You

Yours sincerely,

Narendhiran.R

Interesting course, was curious about what all things will be covered in this course. It touches most of the topics that one should be aware of ML. Only thing that I felt bit overwhelming was the amount of material which was covered in 4 weeks. Could easily be stretched to 5/6 to make it less demanding for a novice person.

Good stuff :) However approaching final assignments I was missing more info about preparation of an input data. As far as I know it is to some extent covered by first course of entire Specialization. So, I plan to take this one as well. But overall - very good intro to ML in my view. Thumbs up University of Michigan :)

Great course, with a very practical overview of the different options available for machine learning models using Python. The concepts are the same as in R-based machine learning, but this course was great for getting experience with which Python functions to use for various machine learning models.

This was a demanding course, requiring one to listen to many long lectures, do a lot of extra reading and research online for more in-depth machine learning knowledge in order to answer all the questions and problems. It was time intensive yet provided a good background on machine learning models.

Great course. It doesn't confuses you very deep mathematics involved in machine learning. Rather, with a touch of it, it focus more on how and when to apply the models in Machine learning. How to evaluate and optimize them. It's really Fantastic with it's hands on projects in assignments.

This course is very cool and interesting. One thing, it would be more useful for me to have a little test/exercise after or in the middle of every video - to try, how I understood the material. Like in Andrew NG course or in Text Mining.

Anyway, thanks for a great course and your work!

Great course, somehow assignments are not always on the same level, the first was easy, the last seemed to be very complex, but was not, the assignment instructions were misleading. Anyway, I enjoyed this course too much and I want now to improve my abilities in underlying theories.

Not only in the last week, all the weeks can have assignments which are like projects. That may give more sense of analyzing and understanding the process of model selection, application of supervised learning techniques. But the course is good, and i have learnt it in faster pace.

kind of a good course. However, I think too much things have been put into this four-week class. All methods, for example, random forest method need a lot of practice. In the four week, I think I am not familiar with most of these method and I need to practice more in the future.

I've learned a lot of basic concepts about common machine learning models and how to apply these tools using python. Although practices and deep understanding are still not enough, this course is really great and worth learning for beginners who want to learn more in this area.