This course is outdated.

too old documents

This course is excellent for anybody new to machine learning and wanting to learn this new skill from the top down. For me, I have a strong background in machine learning, not in the context of big data, but I wanted to get familiar with Python and learn how modern companies are using machine learning in practice. This course provides that applied approach to implementing a broad range of machine learning applications with Python, applied to real problems.

A course this small cannot provide everything - what this course does not provide is in-depth technical tutorials on the workings of machine learning algorithms. There are many courses out there which do, but this course to great for learning a practical approach to problem solving with machine learning and data processing.

If there is a downside, I would say that the use of paid packages in the lectures (graphlab) limits the student's ability to learn Python using the freely available packages on the web, which was my personal preference. However, this is not purely negative, since there are many employers out there who would like to know that you have practical knowledge of things like AWS and graphlab. I did enjoy learning about those packages and services and I feel like I learned something positive which I can share with potential employers.

Overall, a very good concise course - one of the best on Coursera for vocational learning in my opinion.

Excellent overview course. It has exactly the right balance between explaining Machine Learning concepts, and providing enough supporting mathematics & logic to understand why these concepts are correct (without going through epsilon-delta proofs).

Having followed several Machine Learning courses, this is now definitely my favourite new course, replacing Andrew Ng's famous course here on Coursera (which was also very good & especially complete, but required too often a leap of faith - this course provides really more details on the "why"). Furthermore, the exercises in this course are spot-on: they use Python and GraphLab Create (for which you get a 1 year student license when taking this course) - the big advantage is that you can focus on the Machine Learning aspect, and not on how to implement something in Python (or Matlab or R). The exercises are challenging enough and require some thought, exactly what they should do. This is not a "look up the right answer in the slides" course when it comes to exercises, which I particularly like.

The chemistry between the teachers is also very nice and shows they just love Machine Learning, and love teaching it (which they do very well).

If you some familiarity with statistics (a bit) and mathematics (a bit of matrix & vector calculations), and want to understand what Machine Learning is about, then this is THE course for you.

Emily and Carlos, you are the best! Thank you so much for offering this great course. I like your humor, your casual, yet very direct and practical, approach of teaching.

I'm a film student from Germany but I was always interested in Machine Learning and AI - more like a hobby. This course gave me a very good intermediate understanding for the mechanisms behind this hyped and often overcomplicated subject field. The knowledge I gained helped me deliver a way better master theses in film school. I was able to (automatically) collected huge amount of tv-series data on several platforms via import.io and dbpedia and build a really great, combined database (dato's SFrame was very helpful here!). Through the techniques of this course I was able to push the analysis in my thesis a lot further than I ever expected!!! I will try to finish the other courses of this specialization although I'm an expert and professional in a completely different field. It's just so much fun and so comprehensible!

Also I got the impulse for a great sci-fi television series, which I will be writing the next few months now ;)

I came into this course knowing little bout Machine Learning. In fact, besides knowing a touch of HTML, I have no significant background in computer programming. Even before I started watching the first video, I was already expecting this to be an especially challenging course, for me at least. However, I was pleasantly surprised with the content and delivery - Carlos' and Emily's adorably dorky banter and their clear and concise approach to the various case studies made it easy for me to grasp the fundamentals of Machine Learning. Their delivery of the course's content is beyond reproach. (Although I would have loved to see Carlos going on a little more about Messi and soccer in general!). I struggled a little on the last question of the final assignment (Week 6), but besides that, it was smooth sailing. Overall, it was a positive learning experience and I'm happy to say that I now know more about Machine Learning than when I began. If you're new to it, this course is a great way to learn what Machine Learning has to offer.

Excellent content, and at just the right level for a getting-your-feet-wet-course. I especially liked the overall vibe of the lectures, which was relaxed and kind of goofy, and it's actually kind of nice to get some sense of personality from both Carlos and Emily. This is a topic of how to understand and manipulate the world as expressed to you through data -- a completely dry and theoretical approach would be tragic. I eagerly look forward to the rest of the specialization. And I had an ah-ha! moment in the week 5 homework -- it's a fairly simple model of building song recommendations, but when you actually look at the recommendations that come back from this algorithm, you kind of see that it does an intuitively better job than any system you could design and build without using ML techniques. Being a (successful) software engineer, this was both humbling to me and inestimably cool! It's not just a few new tricks to add to my bag-o-tricks, it's a whole new field to digest and investigate.

I'm very excited about this!

A fascinating tour of what's possible today with modern machine learning tools. The beauty and challenge of this course is the approach - diving right in to the tools to work through and experiment with some case studies. This is not a talk and visuals only course. You will be hands on.

This may be demanding for some, but is worth the effort. The course says no previous experience necessary in Python, but I recommend having at least completed a beginner's course before trying to tackle this. (Or familiarized yourself with Python if you have other programming experience - it has its quirks, like every language.)

The course will introduce you to the current state of play in machine learning and both show you what's possible and also where the limitations are. This is not a superficial course (talking points only) - you will learn enough to be dangerous. If you want to be a little safer, do the follow-on courses too. (At this time, only the 2nd course has run - regression - but it was very good).

Presenters start off kind of silly and made me wonder what I was getting into. However this class quickly evolved to be 100 times better than the course offered by U of California on Big Data. You do actual python programming through a lot of serious concepts in data analysis, visualization, and machine learning. This first course is hands on - just use the libraries. They lean heavily towards Dato which is not open source - using a 1 year trial license. However there are better instructions and support for open source in subsequent courses. Also - the second course in the series which I'm taking now is taking what we did in course 1 and diving into the math and algorithms involved - walking through actual proofs etc. It doesn't require you to know them well enough to do on your own, but they do walk you through them and explain extremely well - you actually implement the resulting algorithms. I'm fascinated by this course and can't wait to apply what I've learned.

Very approachable for a beginner trying to learn a few evenings a week. It has a consistent pace and the topics are explained in the right level of detail.

Modules are centered around a Real world problem that is easy to relate to such as product recommendations on a website or analysing text. While other courses dive straight into calculus and theory where it's hard to recall the actual problem being discussed, this course doesn't have any those issues. It is very well structured and gradually progresses while providing real learning along the way.

The assessments are just the right length and offer a suitable challenge without expecting hours of work for each question.

A large benefit of this course is the environment setup is straightforward. I had my jupyter notebook running in a few minutes. With other courses, I was spending hours trying to install things before trying to learning anything but with this course I was up and running quickly

Artificial intelligence been around for long time and machine learning is the application to self learn through the data and apply and predict, be more and more accurate. This was a first encounter for me to know how deep learning and deep feature works! Probably, this was the time when I felt going back to university days and relearn few concept of statistics, in order to understand few prediction model and the usage. I was amazed to see and unaware of the fact, I am benefitting as user and million of users unknowingly. Every field and every industry and most importantly every area of our life is going to improve/ impacted with Machine learning. It is a great effort by the faculties, to bring such complex topics to level where it's looks like story telling and making folks understand through small assignments but surely it is a result of deep thinking and hard work which makes this course so interesting and intuitive.

Machine learning is one of the fields in the modern computing world. A plenty of research has been undertaken to make machines intelligent. Learning is a natural human behavior which has been made an essential aspect of the machines as well. There are various techniques devised for the same. Traditional machine learning algorithms have been applied in many application areas. Researchers have put many efforts to improve the accuracy of that machinelearning algorithms. Another dimension was given thought which leads to deep learning concept. Deep learning is a subset of machine learning. So far few applications of deep learning have been explored. This is definitely going to cater to solving issues in several new application domains, sub-domains using deep learning. A review of these past and future application domains, sub-domains, and applications of machine learning and deep learning are illustrated in this paper.

Loved the material and the course design - it really works for people who don't have much time but want to understand the main principles of machine learning. I think I've watched every week's videos and completed assignments within about 2-4 hours.

The only suggestion I would have (and it is a very personal opinion) is to spend less time on illustrating slightly irrelevant aspects of the material, such as showing quite a few Amazon products or going through a full shoe collection. I can see how that can make the course a little more lively but for a person who treasures every minute of their free time, it can be noticeable, especially when it takes a significant fraction of the very well-sized small videos. This was a very minor issue but I thought I'd share in case someone else felt the same way.

Overall, a huge thank you to Carlos and Emily for a great course!!

The only way these courses could be better if there were far more of them from the same professors. If more of the nitty gritty details of these algorithms were fleshed out in all their glory, more algorithms, more mathematical derivations & more tutorials in the programming languages & libraries used. Otherwise, these MOOCs are near perfection. A very, very nice introduction for beginners with just a little bit of math & not too much programming. Just enough for busy people. I've reserved that 5th star due to the slow pace that the MOOCs have been released (which will presumably be irrelevant for future machine learners) & the fact that there really needs to be more of these very high quality moocs. So there aren't enough of them, so I reserve a star. Hopefully in the future that will be irrelevant as well in which case I'll regret not indicating 5 stars.

This first part of the specialization course provides a gentle, well-balanced introduction to ML concepts. I had some partial knowledge about the subject and some familiarity with the statistical techniques, but the thing I needed most was some real-world requirements and applications. And I am very content with what is presented. The cases are concrete without being complex and while the course is careful not to drown you in theory, it does slowly build an insight into the techniques being used. Former Python experience is recommended, yet even if you have never used Python before, I think you can build the necessary amount of skill during the course. And I must not end without saying a couple of things about the instructors: They are amiable and entertaining as much as they are knowledgeable and that makes it a very pleasant experience.

Overall I thought it was a useful introduction to ML. I actually liked the fact that I needed to install WSL on my Windows laptop to run Turi Create. This provides me access to other programs and compilers that I usually access on Linux servers (e.g., gcc, gfortran, etc). I am fairly familiar with Python, but tend to use Spyder. Becoming more familiar with Jupyter has also been a bonus. One thing that I was not as familiar with was SFrames and the use of dictionaries within Python. I have mainly used Python for numerical linear algrebra, image processing, signal processing, and the like. Links to good tutorials for SFrames would be helpful. As an introductory course, it focuses more on applying techniques rather than specific details related to the techniques. Hopefully, these will be discussed more in-depth in the courses which follow.

Very good course for an overview of Machine learning .

It is very good first step for Beginners. Word of caution : Course lecture videos uses the Graphlab. Course jupyter notebooks have the up-to-date Turicreaatecode. You can also use Panda/Numpty/Matplotlib to arrive to same conclusion. However I am new to ML/Python so stuck to course guided notebooks. it worked for me.

Course first gives you an intuitive and solid theory background and then walks you through how to implement that into Graphlab/turicreate.

Best part is Quiz / Assignment. You are not just copy pasting code from the class but really challenged to find a solution using turicreate API .

Overall good start. Recommend with foot note that be ready to do some googling on Turicreate while following along the course.

Overall, an excellent hands-on course to learn the basics of machine learning. I am really glad I chose this over many other options available online.

If I were to pick straws, a couple of the programming quizzes could be better, especially week 6 images quality and related questions 2 and 3, and couple of the theory quiz questions were misleading.

Finding all the iPython notebooks and data online on Amazon was a Godsend as those data files just refused to upload from my system. I wasn't aware that these were available so I could have saved 2 weeks of delay in completion due to data upload problems.

I am keen to continue my association with Univ of Washington and Coursera based on this experience though I am not in the market for certifications at this stage.

Great introduction to Machine Learning with the case study approach. Gives you a quick preview on what Machine Learning entails iwth practica use cases without making you learn a new language and several frameworks before you write your first line of useful code. The course seems to emphasize heavily of the practical uses of machine learning

The instructors have a pretty fun way of interacting with the viewers/students. Instead of reading power point slides. The quality of the content is great, though it can be made a bit more consistent in a few modules. The labs are pretty great and an important part of the course material.

The format of this specialization should serve as a template for future specializations made on Coursera.

The Course is well structured and gives a good overview of Machine learning and it's various applications. By working with case studies, it is easy to understand the logic behind the application. The only regret is that the usage of GraphLab means that it is a bit difficult to transfer the learning to something more open-source such as sci-kit. Graphlab does most of the algorithm work for you so you might not really understand the backstage working of it. Hopefully the courses further in the specialization will teach what is happening behind the scenes. Also, all the data is cleaned already so it is very easy to work with. you will have to learn data cleaning and data mining separately to be able to use it in your daily work.

Brilliant course!

This course it taught by two amazing professors- Carlos Guestrin and Emily Fox. In the very first session itself, they make it crystal clear that they love what they do, and in the subsequent videos they show you why you should get excited as well.

The course puts remarkable emphasis on the concepts rather than raw code, and once you're prepped on the fundamentals, you're gently introduced into the code. This to me is a wonderful way to go about a new subject.

The slides are simple and clean, the work of a brilliant team shines through the entire time. Carlos and Emily are really funny and you just fall in love with this course because of these two exceptional teachers.

Without a doubt- highly recommended.

Since when I got to know coursera, I have been going through plethora of machine learning courses .I couldn't complete even one of them as the pace was too boring and I couldn't connect anything to the real world scenarios. So I had paid a huge sum and joined a course on classification model using SAS near my place. To my astonishment, this course by the Amazon professors far excelled than what I took up earlier and I feel I had wasted money to pay that institute. This is the best machine learning course ever which has the theory and practical stuff at the same place. Thank you Amazon professors for all the help. Hope I start actively participating in kaggle competitions once I finish the other courses too.

Great introduction.

I have two suggestions:

1- GraphLab is a great tool but it is commercial and it is not available in all companies. Hope there were some guidelines, slides with sample source code etc. for each session showing how to complete assignments with an alternative free library.

2- In the slides, I found very useful practical hints mentioned quickly by the lecturers (e.g., how to fit a better regression model by using the same feature with higher order such as X^2, X^3, X^4...). There is a high chance to overlook such hints since they are mixed with other materials. It would be nice if such practical hints are gathered and published in a separate document like a cheat sheet for future references.

The instructors are amazing and teach at a pace, which is slow enough for things you are new to but fast enough for more familiar things. I had decent machine learning background but none of SFrames. However, I had no difficulty understanding SFrames from the word go. The assignments are the best part. If new methods pop up in the programming exercises, step by steps instructions explain how to use them and you learn even more that way! (some of the hint/help links in the assignments did not work for me). Instructors used a lot of examples during lectures but I felt that theoretical part was lacking depth (may be it is me because I always like to see more maths :D). Carlos was great so were his shoes :D

This whole specialization is an outstanding program: the instructors are entertaining, and they strike the right balance between theory and practice. Even though I consider myself quite literate in statistics and numerical optimization, I learned several new techniques that I was able to directly apply in various part of my job (ok, not in the foundations course, but in subsequent courses). We really go in depth: while other classes I've taken limit themselves to an inventory of available techniques, in this specialization I get to implement key techniques from scratch. Highly, highly recommended.

FYI: the Python level required is really minimal, and the total time commitment is around 4 hours per week.