Most fun: Pat Pattison, Songwriting, Coursera. Very good lectures, very good material, very well presented. Teaches a lot about writing song lyrics in just 6 weeks, breaks it nicely down to steps and recipes. I used to think that the best feature of MOOCs is the automatic grading and feedback from programming homework, but in this course, for the homework songwriting you gave and got feedback from 3-5 random people in the course, and it was not only useful but this feeling of togetherness with strangers was even better than getting instantaneous feedback from a bot for programming homework. Shows that teaching art scales to MOOCs as well.
Nicest: Andrew Ng, Machine Learning, Coursera. Interesting topic, well-planned material, very well avoids going into the mathy details, but still conveys a feeling of understanding of the topic, so accessible to a wide audience. (Martin Odersky, Functional Programming Principles in Scala, Coursera, was almost equally nice, but had some rough edges in the first run.)
Most interesting: Probabilistic Graphical Models, Daphne Koller, Coursera. Very interesting topic. I took the first run of the course and it had lots of rough edges. Needs a lot of work to apply the lectures to the homework. I haven't seen such a demanding course since I took quantum mechanics at university.
Best organized: Jennifer Widom, Databases, Stanford. This is not the flashiest of a topic, but oh boy was it well organized. Runs like a clockwork. Everything in the lectures is relevant, everything from the lectures is applied and tested in the homework, there is lots of homework (but still not enough to make you remember SQL,XPath,XQuery,XSLT for the rest of your life if you don't keep using them), weekly homework has a nice progression from simpler things to medium difficult things, and the web environment is well designed, and gives wonderful feedback and guides you to get your queries correct.
> Nicest: Andrew Ng, Machine Learning, Coursera. Interesting topic, well-planned material, very well avoids going into the mathy details, but still conveys a feeling of understanding of the topic, so accessible to a wide audience. (Martin Odersky, Functional Programming Principles in Scala, Coursera, was almost equally nice, but had some rough edges in the first run.)
Have to agree with all of this. I've taken Andrew Ng's Machine Learning course (only time I've paid for a 'verified' certificate), and found it a great overview of ML, though I'm not sure I'd feel comfortable telling anyone I have a good understanding of ML :)
Odersky's FP in Scala was actually the first Coursera course I took (during its initial run, I think). -- I also found the follow up Reactive Programming course to be excellent as well.
I too loved Odersky's course. It was definitely rough, but having the creator of a language teach a course about it provided insights that I wouldn't have gotten otherwise.
The Odersky's course is phenomenal. Highly recommended and it shows how much attention and craftmanship was put into designing Scala. Bonus: Martin speaks like Arnold and it's very enjoyable to have a "Terminator" voice teaching you complicated material.
Interestingly, you took the same path I did with Scala and ML. My criticism about these courses is that some of the projects and content can be too easy to get right, skimming on the surface in some areas that would need more time to grok. Lately I've moved to Udacity and there I can find more in-depth projects and discussions with virtual classmates. The price is steep but you pay for what you are getting.
Agreed Ng's Machine Learning and Odersky's FP in Scala were my favorites. I'm looking for a good bioinformatics course at the moment. I wrote a small program for my daughter that attempts to find CRISPR sites for my daughter and it would be great to know more of the background.
I wish I enjoyed Andrew Ng's Deep Learning courses. For him being the cofounder of Coursera the production quality (audio/video) was pretty lacking. The whine/distortion on the audio made it difficult to listen to on headphones. Many of the exercises were either a bit too railsroaded and simple, or poorly explained in their goals and barely worked. I really didn't like his style of writing on the digital whiteboard. Perhaps just a side effect of a MOOC, but he never checks for understanding in even the most basic of ways.
While I'm not perfect, I spent nearly 3 years teaching at a bootcamp fulltime, so perhaps I have different standards for communicating and teaching actionable lessons?
> Jennifer Widom, Databases, Stanford. This is not the flashiest of a topic, but oh boy was it well organized.
Couldn't agree more. I took this course in 2011 but didn't have a need for working with databases until 2014. Three years after I took the course I was able to jump in and work fluently on databases -- Mongo, Sqlite, Firebase, etc. The least I can say is the course helped me internalize database concepts.
This course has the right amount of handholding yet challenges you enough so you acquire long-lasting skills.
Ditto. I was able to get my foot in the door of industry due to this course and Scott Allen's PluralSight C# courses.
I've now built 2 data warehouses and helped with maintenance on another. It's not my main focus, but it's nice to be able to do it myself when the work calls for it.
I got hired at my first job out of college due to the C# ones and I've been working in C# since then, with a smattering of other languages here or there.
I studied portions of it before I got my first "real" tech job (not a call-center) and it put me on a path straight towards my career as a developer. I go back to the course every year or so and try to pick more gems out of it -- it's truly fantastic.
That was the first MOOC I took as well. It turned out to be an excellent entry point for databases. Perhaps because it was one of the first Stanford MOOCs, the quality of it was high.
This thread's full of really valuable information, so I've compiled these and a bunch of the other courses mentioned here into a big document[1]. I've also added quotes from this thread, with links back to the original comments.
Andrew Ng's class was great! For a tougher class that focuses on one of the technologies, I recommend Geoffrey Hinton's "Neural Networks for Machine Learning" on Coursera. Really eye opening for me, and fairly close to the leading edge in deep learning, as far as I can tell. I felt that the exercises were more detailed and challenging than in Ng's class, and thus I ended up learning more.
– the first course in the specialization has a very good and engaging start, but the gap between lectures and problems widens quickly after that (maybe that's why the author boasts about a "challenging" course "not for everyone"). I'm hesitant to take the next course of the specialization.
MIT had a similar course on edX and that one was brutal as well (Computational Probability and Inference). I guess nobody figured out how to teach it the easy way.
I've been meaning to either take the course or read the book. I'm curious if you've read the book [0] and how you would compare or if you'd recommend one, the other or both.
For me Bayesian networks are a tool of thought and I think it's worthy to learn them in the same way it's worthy to learn to sketch functions with pen and paper.
Yes, I think so. PGM or some improvement on that is relevant to doing things like reasoning correctly based on evidence.
I envision more sophisticated AGI systems to use DNN or other NN techniques to learn about the world, and be able to take in uncertain input and make sense of it. PGM or similar would then be used to correctly (in the mathematical sense) reason about what to do to accomplish the agent's goals.
I too took the first iteration, and it was... quite terrible.
The lectures were OK, but the homework was more than challenging. Not only had you to battle with the topic itself, but then you need to magically acquire knowledge in some totally unknown topic (I think it was genetics) and wrangle the quite baroque representation of that topic shoehorned in a programming language that is totally not made for it.
I was on the verge of despair because of those secondary problems. Really.
It's good to hear that this MOOC is still so well thought of since I first took it; for me, it was the first course I took that made me really understand how a neural network and back prop actually worked.
When I took the course, it was in 2011 - and was known as "ML Class"; yep - I was among the first "beta tester guinea pigs" of the course. It was fun and amazing to participate in.
One of the early participants was even inspired to replicate CMU's ALVINN self-driving vehicle in miniature:
Pat Pattison, Songwriting, Coursera! I took this too! Fantastic! Delightful to learn something (anything!) from someone who so thoroughly and completely knows just what he has to say to teach a topic he's expert at.
Yes, I enjoyed the Pattison songwriting class as well. At first his applying the "strong" and "weak" concepts to every aspect of a song lyric seems overly simplistic, but it actually starts to make sense. I came into class with considerably previous experience in writing poetry, but thought I learned a lot from him and my peers. His breaking down a performance coaching example was also very instructive.
Other music MOOCs I enjoyed:
The Berklee "Developing your Musicianship" series on Coursera taught by George W. Russell. Started off thinking this was too elementary, but the ear training is valuable, and I learned a lot about the use of diatonic chords, and even the few simple patterns he taught improved my song writing enormously.
The Berklee "Jazz Improvisation" class taught by Gary Burton. Very cool to be taught by a living legend, and his selection of songs was refreshingly modern. On the down side, skill levels of the students varied widely, so peer review was more miss than hit.
> The Berklee "Jazz Improvisation" class taught by Gary Burton. [...] On the down side, skill levels of the students varied widely, so peer review was more miss than hit.
100% miss for me. The main thing I learned from that course is that a "MOOC" relying on peer review for feedback is a colossal waste of time, and should have stayed as a video lecture series.
True, I did not learn much from the feedback. On the other hand, the existence of peer review might have pressured me into putting more effort into my exercises.
On the Pattison one, out of curiosity did it use his Writing Better Lyrics for the material? I've read the book and love it to pieces even though I write prose, the way he explores words and phrases is sort of magical.
Nicest: Andrew Ng, Machine Learning, Coursera. Interesting topic, well-planned material, very well avoids going into the mathy details, but still conveys a feeling of understanding of the topic, so accessible to a wide audience. (Martin Odersky, Functional Programming Principles in Scala, Coursera, was almost equally nice, but had some rough edges in the first run.)
Most interesting: Probabilistic Graphical Models, Daphne Koller, Coursera. Very interesting topic. I took the first run of the course and it had lots of rough edges. Needs a lot of work to apply the lectures to the homework. I haven't seen such a demanding course since I took quantum mechanics at university.
Best organized: Jennifer Widom, Databases, Stanford. This is not the flashiest of a topic, but oh boy was it well organized. Runs like a clockwork. Everything in the lectures is relevant, everything from the lectures is applied and tested in the homework, there is lots of homework (but still not enough to make you remember SQL,XPath,XQuery,XSLT for the rest of your life if you don't keep using them), weekly homework has a nice progression from simpler things to medium difficult things, and the web environment is well designed, and gives wonderful feedback and guides you to get your queries correct.