ミシガン大学（University of Michigan） による Applied Machine Learning in Python の受講者のレビューおよびフィードバック

I thank Dr. Kevyn Collins-Thompson and Coursera team for the excellent course. I have learned valuable skills from the course. Dr. Thompson explained ML concepts very skillfully and made the course fun to follow. Assignments are very well selected and reinforce the class concepts. Over-all the course encourages learner to investigate and apply different ways to do same task. I recommend this course to those who are willing to learn machine learning and can't decide where to start.

A solid course. The help found in the forums was also way more useful than the first course in this series. While course two was generically useful, this third course was technically useful. A very good introduction into sklearn. The video instructor/professor was also very clear and methodical in presentation. The assistance by the class monitors was leaps and bounds more useful in this course than course one (I almost quit after course one because of it, so glad I didn't!)

excellent course for following reasons:

1. Excellent i python note books. What ever a student must know is kept in it.

2. every topic is explained simply and well upto what ever we need to know.

3. if you are not in academic field(not planning to do phd on this stuff). Trust me how ever advanced courses you do but after a week or month. these are the points which one need to remember.

4. Course and programming labs are in perfect sync.

Thank you very much for keeping this course

Great intro course to Machine Learning. Gives you a good overview of the main models and Python needed to code. I liked the fact that it did not get too detailed into the Math foundations of ML. There are other courses for that.

I can apply what I have learnt right away on my job.

Highly recommend.

One Note: this course is over 2 years old and the Staff is pretty slow to respond. But the Forums have enough information to get you to self-solve your problem.

Good luck.

Great course filled with a lot of details. The course does a great job in teaching all the important concepts. I felt the feature engineering should have been a dedicated topic. I got a lot of hints from the discussion forum and surprisingly there are even more concepts you have to learn for building a pipeline, treating categorical and numeric features differently. Overall challenging week4 assignment gives you confidence to deal with real world problem.

I have done several of Coursera Courses and also from Udacity (Deep Learning Nanodegree) and I find the courses from the University of Michigan really good. This one for Machine Learning is really specialized for the Application of Machine Learning Algorithms. Sometimes a little too superficial, but it is enough for start working with Machine Learning. The Test at the end of the week are a little difficult but you learn alot from them :-)

I think there were some small problems in the assignments and quizes but all in all those problems made this course assignments even more powerful because it demanded more effort to answer those questions properly.

Totally if you want to get a good sense of machine learning and step into AI , this course will not only give you basics and principals but also you will be able to build and understand different models using python.

good luck!

All the exercises were very challenging and allowed me to apply all the knowledge acquired during the lectures and even more. I loved the fact one has to search for extra information for doing the exercises, because that pushes me forward to learn to search in other sources. Also loved the freedom that there is when solving the final assignment. That is the best expression of a real world challenge and allowed to exploit my creativity.

This course gave me some tools to use in real life. It's pretty abridged in time because they are trying to cover a very big topic in only 4 weeks. It won't give you a comprehensive set of knowleadges, but a good basis to proceed by yourself. Also some basic knowledges are reqired in computational mathematics, statistics and programming for applying this course. I highly recommend this course as a first step into machine learning.

One of the best ML courses on the platform. I highly recommend it to all data-science enthusiasts. It would be nice to have pandas data-wrangling skills before tackling the final project as it is a must. Totally enjoyed the final project! was a great learning experience seeing my classifier AUC going from 57 all the way to more than 76 and the impact of feature importance and cleaning on the model performance was eye-opener!

Great class! I had fun learning many new things in this course. The professor did a very good job at taking a complex subject and making it simple and easy to understand. The code and assignments were straightforward and not overly difficult. The real quizzes/tests in this course were appreciated as this felt more like a "real class" where one can really learn a lot. One of the best online classes that I have taken.

Very useful and true to the name, it teaches Applied Machine Learning - how and when to carry out the various algorithms on a dataset, how to tweak the parameters and tune the model. Really Really helpful if you're looking to finally get your hands dirty on data after reading all that theory!

Also gives brief but necessary summary to all the different algorithms with intro to deep learning as well. Highly recommended!

I thought this was a very good course in Machine Learning using Python. I took Andrew Ng's Machine Learning course before this one, which I would highly recommend! I enjoyed this course because it taught me about scikit-learn, which I plan to use in my career. I also purchased the recommended textbook "Introduction to Machine Learning with Python" from O'Reilly, which I found to be a very useful reference.

The course is well done and both the lectures and the practical assignments have generally a high quality. If you come from a theoretical background, be aware that this is a very "high level" course, meaning that a lot of attention is put on the practical application of the different ML methods (using the sci-kit learn library in python), but very little is said about their mathematical foundations.

All complicated math acknowledges were cut off and fully focused on applying ML using python. As an energy engineering master student who doesn't have much programming experience, I find this course very useful. PS. I've previously taken the specialization 'Python for Everybody' to get familiar with python. I suggest doing the same if you also have no idea of python just like I did when I started.

Excellent instruction and challenging assignments! Sophie from the teaching staff was very helpful and responsive to forum posts. Thanks to Kevyn Collins-Thompson for a great survey course in machine learning. The only downside was that the auto grader has limitations which inhibited some exploration (one can not keep plots in the submission is an example), but I'm sure that will get worked out.

The course is a great overview of the basic algorithms that every machine learning practitioner should know. Since it has a limited amount weeks to cover such a broad subject, you will have to dig a little deeper by yourself. I found the reading material also very interesting. The final project is awesome and it will definitely make you experiment what is exactly what a Data Scientist should do.

This is a great course I learned a lot, especially it familiarize me with the SKlearn toolkit which is very very handy. I notice that the SKlearn documentation contains a good figure which shows a rule of thumb which learner to use. I recommend you to include in course reading, because some students might find it very useful.

http://scikit-learn.org/stable/tutorial/machine_learning_map/index.html

This was an awesome and engaging course. Machine Learning is a vast field with lots of ground to cover. This course gives a broad overview of all the different parts of machine learning without going too deep and also keeping everyone engaged. The assignments, especially the last one test what you learned and keeps you on your toes. A good beginner course to Machine Learning. Thank You!

Very informative about machine learning approaches ie supervised and unsupervised learning. And then goes into detail about the techniques such as regression and classification for supervised learning and clustering (K-Means) for unsupervised learning. Other techniques are discussed such as Principal Component Analysis etc.

I enjoyed it and would recommend for all data enthusiast.

Kevyn is an absolute joy to learn from. His enthusiasm for the topic is contagious, and his explanations are clear. The course content is well curated, tested, and reinforced. At the end of this course I feel confident that I can *actually* apply machine learning to real world problems and competitions. This is not just a 'good' course, it's a new gold standard in e-learning.

Much more detailed than the previous two courses. The lecturer teaches with more verbose slides and thus gives you a more detailed overview than the lecturer in the first two courses in this specialisation. The assignments are much easier as well. But still thoroughly useful and I have to admit a welcome break from the gruelling process that typified the first two courses!

The course allows you to play with a multitude of supervised (and unsupervised - optional) machine learning methods. Professor Collins-Thompson is very clear and knows perfectly well how to convey concepts and how they should be applied in real situations. I recommend it to anyone like me who needs to be able to develop simple codes and understand what they are doing.

Very good course for a wide spectrum of audience interested in Machine Learning. I just had a basic learning of ML and Python, but the course was structured so well that I could catch-up. Also offers an interesting peak into Neural Networks and Deep learning. Overall, an excellent course with clear and attainable objectives, backed by high quality content and data.

This is great course with very practical methods to sovle real problems in various fields. I think there should be a additional course regarding Deep learning, which I think would be very successful as well.

Moreover, this course can be combined with Andrew`s ML so that we can have both theoritical concepts and practical experience of Machine Learning in python.