A great tutorial on using machine learning tools and a very basic intro into machine learning concepts. Doesn't go into how they work at all which is what the other courses are for.

Given its basically a tutorial for graphlab create (awesome but commercial software) it would be nice if it also covered pandas directly also. If people are really expecting to head into this field they are likely at some point going to have to use the cheaper tools.

It would be great if this module was optional for the larger machine learning certificate. That said I found this class was valuable in preparing me for the next courses in that it got me familiar with the tools and the rhythm of the quizes and assignments. Someone more studious or smarter than me may find this course less valuable.

A great course , I loved the combination of theory and practice , the study cases really helped cement the understanding. Both professors were coherent , didactic and helpful in advancing my understanding of machine learning. Their quirky, nerdy humor was endearing, their knowledge noticeable and impressive.

I'm definitely going to continue auditing the rest of the courses in the specialization. I'm still trying to use graphlab as much as possible, SFrames are so similar to pandas dataframe that it's not hard to do , however , plotting more complicated graphs than a scatter or line is still sometimes a challenge.

I strongly recommend this course to anyone looking to explore and strengthen machine learning concepts in a vey applied manner. Congrats to the team at UW

Overall, I really love the course and the ways the instructor explains about the subject in a very practical manner. I have a lot of useful skills in machine learning. However, some of the excises wordings and instructions are confusing. All instruction should be on Jupyter Notebook which makes it easier to read and follow the instruction right away while working on the project. Some typos in the assignments make data retrieval very difficult and frustrating ( Kings of Leon in the assignment, while the correct name in the database is Kings Of Leon etc.-a a few other names were not correct like George W. Bush). General software set-up should be clearer since to have "turicreate" to work on Windows was a pain.

This course gave me the literal basic understanding of machine learning. It essentially cant get easier than this. It gave me the required knowledge to start of and then explore the world of AI on my own. I started this course with no prior knowledge about ML. Although this gave me a basic understanding, it didn't into the right details, in the sense that it never gave a good understanding of the math behind a lot of the algorithms, nor did it give any insights into how to build any of them from scratch.But then again, this is a bare bone beginner course not meant to house information such as those.All the details regarding the math and building one on my own, i used outside sources to learn along the way.

In a fairly short while (3-4 days of full time study) this course provides a good and basic overview of fundamental concepts of machine learning.

I have an M.Sc.EE from '91 and was able to follow the course content and exercises without any problems.

The majority of my time was spent in Python doing the exercises, and doing that, I got a better understanding.

The video sessions was well produces, quite interesting and (mostly) to the point. Carlos' and Emily's good mood and energy makes it easier to keep focus.

However, don't expect to become an independent master after this course. It's just an appetizer allowing you to understand what's being talked about. But that's a good start, right?

The course was overall good in achieving its goal of trying to introduce various machine learning methods, without getting into too much detail. The course used IPython Notebooks, which I found to be a very useful environment for learning. The professors have done an excellent job of presenting the materials with good examples and code.

I took a star away because some of the quiz questions did not stimulate my thought process much and relied on remembering terms and definitions. Also, I would have preferred the use of open source Python libraries as much as possible and use GraphLab Create only when using open source libraries would have been very cumbersome or impossible.

This course provides a broad overview of many of the topics in machine learning such as regression, classification, deep learning, and recommender systems. The lectures are well done and interesting and the Jupyter Notebooks provide walkthroughs of the concepts covered. The only negative is that this course is not rigorous enough. There is very little in-depth programming and all of the assignments focus on a high-level introduction rather than a rigorous implementation. Overall, a good introduction to the concepts, but this course alone will not provide learners with the skills needed to build a novel implementation of any of the covered machine learning techniques.

I thoroughly enjoyed the course and discussion forum was very helpful to progress. However few improvements is required to have better user experience 1) The content should be upgraded to latest python 2) Since the course recommends Graphlab, there should be up to date detailed instructions to install the same ( disclaimer that this course is not about software....is not helpful to progress) 3) No response is discussion forum - there's no responses provided to the queries. I do understand many questions in discussion forum is repetitive and someone may find the answer by scrolling through long unstructured hundreds of responses, however this is not efficient.

This course is the first of the ML specialization series and provides an introductory overview of the content. I do like the case study approach which allows a grasp on the real application of ML methods. The only part I found a little vague is the concept of whether a method is "personalized" or "can capture context" (in week 5, I think).

Since this is the only introduction part, the details are deferred to later courses which probably matters more. I have finished the next one (regression), and I do find the material there more in depth and informative. So if you feel the content in this course a bit superficial, you probably still want to try out the next.

This course is a pretty cool intro to the GraphLab Create library. However, I would have preferred something more challenging, even for a survey course.

Though the discussion forums were generally very active (I posted at least one question that was answered by a TA very promptly), there was a question I had towards the end of the course that never got answered. I wasn't alone in this... many people were having a very similar problem.

In general, I think the course was good. But, if you have even a very shallow familiarity with Machine Learning and/or Python, you may be able to skip this one and move straight into the Regression course.

This course is good enough for introductory level, and knowledge. However keep in mind that:

If you are new to Python, you are in for a quite big a challenge. The course are kind enough to walk you through machine learning steps, BUT will not hold your hands through Python programming.

The course focus on using GraphLab Create, which is free for academic use, NOT for commercial. This look like a bad decision to choose for a MOOC. However, when you find out that one of the instructor is CEO of Turi, things start to come clear.

Also, staff are not responsive at all on the discussion forum. The only active one are the students and mentors.

The course is really good and would I highly recommend to others. This course builds upon your machine learning foundations really nicely. But I would say, anyone who takes this course must take other courses from this specialization as in this course we used only the existing functions for all he machine learning algorithms. So in order to implement those algorithm yourselves, the further course must be taken.

One more thing, there is a lot to learn in this particular course, which is a good thing, but can be heavy, specially if you're trying to complete the course in a short span of time (I had to do this in 2 weeks).

Loved that class. My only issue is that it uses GraphLab (which is Prof. Carlos Guestrin started, so there is some potential bias on the pertinence of that library) instead of a truly open source library like scikit-learn. As a consequence you need to acquire a license for GraphLab, which is a little tedious (and will expire). Also, their notebook system doesn't work very well (at least for me) and you learn ML on a platform that you will need to acquire a license for in the future (it is unclear how much it is).

That said, the material is very well presented, clear and exciting. Looking forward to dig deeper into it.

Fantastic course. Carlos and Emily are amazing teachers. The let down i had was that there were no proper instructions on downloading graphlab, jupyter notebook etc. Due to lack of guidance , i spent majority of my first week fussing about, downloading course requirements. I wasted a lot of time in downloading stuff rather than enjoying the course. Hence, i would recommend the course creators to include a small segment on how to download the requirements and into which file. Otherwise , this is a highly recommended course and the case study approach makes it more practical and relatable.

If you have a computer with Windows be prepared to the challenge of installing Turi Create package which is recommended in this course. Videos are not updated and use Graphlab, while all text instructions tell you to use Turi Create. However Turi Create only works with Mac or Linux. So you have to install Linux within your WIndows with WSL. It's a real trouble and not described in the text.

Also there will many errors you'll be getting when using commands for older packages (this commands in the course are just not updated).

Overall the course is good especially videos and speakers.

The course was great in terms of the content and Emily and Carlos were amazing Professors as I can see how passionate they are about Machine Learning. However, I am giving this course a four star because some of the materials are really outdated and does not fit with the current version of the software which makes it hard to connect sometimes. For example, in one of the quiz questions, the "correct" answer was actually not correct when using the data provided but it has to be selected to pass the quiz. I'd love to see a more updated/recent version of the materials covered.

This is a very awesomest course for novice learners. 👌. A little bit outdated but nothing an hour of google search will not solve. Use of SFrame was a bit confusing. I mean, I understood the difference between SFrame and Turicreate after many many lectures. it should have been explained in the first few lectures. And there is another confusion about the codes of graphlab which the videos contain while you code in Turicreate. It is hard and this is exactly why it is the best way to deeply understand the concepts. Going to study other courses in the specialization as well.

The course, as intended was just an overview, which sometimes can leave the student with a lot of questions. (which i guess was intentional too). Also i feel the instructors were trying a bit too hard to sound 'cool' and 'fun', or maybe it's just that i am 32 and i cannot relate to it much. The content was excellent and the course serves an ideal platform for the launch of the specialisation.

Kindly give other free options of software, since graphlab is paid. I small session on scikit, nolearn or any equivalent machine learning package would be extremely helpful.

I really enjoyed this course and found it fun to use the iPython notebook to play around with the ML models. The instructors were great, but I encountered some issues in the assignments for the last two weeks where the understanding that was provided in the lectures wasn't enough to pass the assignment. Since all the machine learning models are used as a black box, I think it would be beneficial to audit this course and then pay for the others in the specialization. Nevertheless, I truly feel like I have a good high level understanding of Machine Learning.

A very good introduction to Machine Learning Concepts & Principles. Easy to understand, while at the same time, giving a solid foundation on the models.

The only detractor is: usage of GraphLab - feels very custom and thrust upon. Definitely easier to use due to the abstraction inherent ion the programming constructs provided but IMHO, would have been much better to use more generic and popular packages like sci-kit, TensorFlow, etc.

Finally, the DL concepts in week6 felt a bit rushed and confusing. Admittedly, it maybe hard to abstract into a week.

Pros:

a) A good overview on machine learning, with case studies that are practical and can really help you remember what machine learning is about.

b) It requires Python, which is a language that more are familiar with than R.

Cons:

a) It uses a proprietary (and expensive) Python module. It's free for academic use, but if you want to go into data science startups, using GraphLab may not be as practical, as, say, pandas.

b) If you want a more detailed discussion, go for the next courses, or read an ML textbook.

Overall, I'm happy with what I learned.

Good introductory course. Provides hands on experience with toy examples that helps to better understand the content. Light on math so should appeal to a wider base. I completed the course in a couple of days. Lecturers are very good with clear exposition of material in most modules. The weakest part was discussion of deep learning but it is hard to fault the course providers since their aim was to provide an overview of the field. Really looking forward to rest of the courses in the specialization to gain more in depth understanding.

Loved the speakers enthusiasm and the breadth of ML applications and topics presented in this course. I particularly liked the practical approach it is base upon. I was a little bit disapointed by not using Open-Source libraries like scikit-learn, pandas, etc. (but this seems a recurrent comment made by others. I also spent quite some time struggling with various turi related issues (license, expired certificates,, etc.). Finally I regret the last 2 courses (5 and 6) were removed from the specialisation (and from coursera as well).

Pros: Good course organization, very good explanations, a nice introduction to basic concepts of Machine Learning, because the entire specialization is created by the same instructions it has good continuity and there is a sense of logical progression between courses.

Cons: The instructions on the assignments can be a little confusing for absolute beginners (but the forums are a very good source for help), the license to use the GraphLab Create framework for free for this entire specialization expires after just a year.

It was a great experience while having this course I learned a lot of innovative modules. The course was swift was not too rapid and steady I would definitely recommend this course for the aspirant who is looking for a head start in the machine learning career. I might have a suggestion for the mentors that the course python notebook are not up to date and tools used to explain in the videos varies with the python notebook this causes chaos in the user's mind and the user might get disturbed or distracted