Hacker News new | comments | show | ask | jobs | submit login
List of the Most Popular MOOCs (onlinecoursereport.com)
200 points by Schiphol on Oct 26, 2015 | hide | past | web | favorite | 85 comments

Im Noticing a lot of these courses are coursera, and one of the most frustrating things about the plattform is the fact that most courses close and access to course materials/videos etc. is revoked after the course is over, this destroys the whole point of a mooc for me, and this choice seems arbitrary.

Why cant education materials just be open and remain open to everyone, journals and textbooks included? There are enough tax paying citizens all over the world pumping money into our education systems that there is no reason it should not be freely available 24/7 to all.

The fact that unlimited free education still doesnt exist in 2015 makes me sad everytime I think about it.

Andrew Ng is one of Coursera's founders. He's big on learning as the development of habits...in American educational philosophy this idea can be traced back at least to John Dewey with academic formality. So it's not surprising that Coursera often chooses fixed schedules and cohesive cohorts over self-pacing. Indeed, Ng's most popular course, Machine Learning is open enrollment, but not self paced and sends reminders when the student falls behind schedule.

That said, there are plenty of courses on Coursera where the materials stay up and online for years. But there's nothing wrong with teachers controlling how, when and where their course materials are accessible, and since many of the courses are from private institutions, e.g. notably Ng's Stanford, tax paying and citizenship (two orthogonal concepts orthogonal to accessibility of course materials) are mostly tangential to major MOOC mechanics.

It's absolutely amazing that the resources are so freely available and high quality. I am not sure where one would get standing to make demands upon the private entities providing them.

I heartily recommend using this[0] tool for any Coursera course you are currently taking.

[0] https://github.com/coursera-dl/coursera-dl

I agree. Coursera's approach seems to capture the worst of both worlds of traditional and online ed. It lacks the live lecture and person-to-person contact of a traditional college course. Yet, it maintains the timeboxed approach of traditional ed, which exists chiefly to provide simultaneity, to scale real-life human contact to lecture-hall size.

Now I realize that some aspects of the way some courses are run require simultaneity. Things like human-powered grading, by TAs or peers. But they don't have to be designed that way. I've taken Coursera courses that are largely machine graded. For those, there doesn't seem to be any real reason why the courses are timeboxed. And even for courses where this is difficult to design around, it seems like it should be possible to tighten the iteration cycle.

> ... worst of both worlds ... timeboxed approach

Just to provide the other side of the story, I've done about a dozen MOOCs (mostly through Coursera) and I'm most of the way through Georgia Tech's online masters program.

Having to make progress every week is contributes a lot to me actually finishing these courses. The Udacity courses I've signed up for never get touched because there's no urgency. Ditto the textbooks on my shelf that I keep meaning to go through, and the Open Yale and MIT OCW courses I've bookmarked... I obviously have the time, but unless there's a threat of "Do it now or it's gone" I'll probably go and watch TV when I get home from work.

I think that's fine, but if the urgency is manufactured, I'd argue that it should be an additional feature to provide that urgency, rather than the default.

I'm the other side of the coin. I may start off on schedule, but once life intervenes and I fall behind, and I've gotten a couple zeros, I'm like why bother catching up? As a result, I've only successfully finished one out of the 6 or 7 Coursera courses I've started.

I've taken a number of Coursera courses. The format works better for some courses than others. The key is that there is a recognizable student cohort and that the students are enthusiastically engaged in live discussions. That tends to happen less in the fifth offering when the primary instructor has moved on to other things.

Anyway, there's pedagogical theory that favors time boxing [and some that doesn't]. Coursera is opinionated in the matter. That's not a controversial opinion in higher education circles upon which Coursera styles itself and from which it draws it's educators.

I'd be interested in reading about that theory, if you have any good resources. I assumed it was for pragmatic reasons. I wonder if the theories consider a context of the opportunities the Internet would create, or if they assume an offline context.

I'm not sure "live lecture" is a unique positive of a traditional college course, at least if by live lecture you mean room of multi-hundred people. In fact, I'd argue that this aspect is one thing that MOOCs can handle very well. Of course, it's also something that remote learning has been able to handle well at least since the VCR was created. I'd argue that the other thing that MOOCs can do well is automated grading when that's appropriate for the material.

What MOOCs do very badly is peer interactions whether for discussions or grading. Part of it's scale. Part of it's the vast differences in the levels and expectations of the participants. Of course, as you say, it's precisely for these reasons that we force courses to provide simultaneity.

That is a big reason why Coursera is so appealing and it won in the MOOC race. By having a few open sessions a year, they can provide official support from course TA and students all gather in an active forum to discuss the material and help each other. It's why MIT's OpenCourseWare is not nearly as popular as Coursera. I'd argue it's a must for the to keep going with this approach, otherwise it will turn into another OpenCourseWare.

There's really no "right" answer to how asynchronous courses should be. On the one hand, if you just open things up, make some videos available, maybe some machinery for auto-grading where appropriate, I'm not sure you can meaningfully call that material a MOOC at that point. It's some YouTube videos with an associated website of which there are already many for all sorts of content.

On the other hand, if you run things on a tight schedule, real life will get in the way for a lot of people in that many of the better courses are fairly significant time commitments.

>There's really no "right" answer to how asynchronous courses should be.

Sure, but given Coursera's popularity it seems to be doing something right.

>On the other hand, if you run things on a tight schedule, real life will get in the way for a lot of people in that many of the better courses are fairly significant time commitments.

But is it? The time commitments are watching few videos, learning and doing the assignments, all of this in a week. I think it's enough even for people working full time. If you don't have enough time on your hands to do all this, I'd argue you don't have the time required to complete the course.

Furthermore having time constraints is very important for students to keep on track.

I don't really disagree. As I said, if you take away all the constraints, what's unique about Coursera or any other MOOC, really?

That said, people with busy schedules, travel, etc. can easily get behind on a more time-consuming course and, at that point, the natural reaction is just to drop it.

My personal preference is to have structure but to build sufficient wiggle-room in that someone can catch up a week or two.

Can't remember exactly which courses, but there were some which dropped off your lowest assignment grade. Of course one could argue that it isn't really about the assignment, but still psychologically, I felt that was a good compromise between maintaining the balance between flexibility and continuity.

One of the things I've always found a bit odd about Coursera is the lack of general subject forums. Sure, you can meet with and interact with some students in the forum of a particular course, but then when that course is over the forum empties out. And if you are taking the course on your own schedule, the place is often empty because you're doing a lot of the work after the course ended.

I understand that open ended courses might not have the volume to create an active forum, but why not create subject forums where people can discuss things and create a community? In my experience, communities are a very important way to give people the motivation to complete some of these courses.

I eventually settled on using edX exclusively[1], which I don't anticipate changing except for specific advanced courses on the other platforms.

My intuition is that the venture-backed model of the other platforms will make it extremely difficult to maintain a long-term vision of free, ubiquitous education. I did find one edX course which decayed, simply because it required student journaling and deeply personal discussion, which would have been inappropriate to open to the public.

1. Not as exclusively as I thought - I just saw a great course from Coursera and signed up. To date I've only completed edX courses though.

It may depend on the copyright agreement Coursera made with the instructor and institution. Understandably many course creators may not wish their course materials to be consumed indefinitely for free without themselves ever receiving any compensation.

That said, I agree with your main point, which is that if the billions that go into the journal and textbook scams were redirected to instructors (via higher salaries and resources) and students (via lower tuition costs), we would have a much stronger education system.

I did a maths course with Coursera that finished early this year. I've just checked and I can still access the course materials [1]. This is a course that is still running new sessions, and that may not be true for courses that end.

And its worth remembering that Coursera allows videos and pdfs to be downloaded, which I get into the habit of doing as I progress through a course.

[1] My Courses -> Archived -> Course Page -> Select session that I did -> Go To Course.

I remember the same thing, but it may be a case by case option. Many coursera courses had subtle differences (grading rules and process).

Yes, I think it is a case by case thing. The stanford courses are always available Im pretty sure, I know others fall out though. Also the fact you cant just click and download the whole course as a compressed file (like MIT) is a shame, especially for people with limited or no internet connections like where my family lives. There you really do need to be able to go to a library and download the course which you can then take home for your studies. People in tech forget this but believe it or not over half the world doesnt have internet (http://www.internetworldstats.com/stats.htm).

The problem is that this only works if you where signed up for the course when it was active. If you find an interesting course in their archives that has already closed, you often cannot look at the course material.

True, I guess. But then again Coursera isn't a library.

As other commenbters have said, course material can often be found on the net if you look in the right places.

It looks like Coursera has recently opened up a lot of classes that had been closed.

For example, Hinton's class on neural networks from 2012 seems to be open again. https://www.coursera.org/course/neuralnets

I believe that Coursera courses are popular exactly because they only allow watching archives if you have signed up before. What I mean is people sign up to a large number of courses, just for being able to view the course any time in the future.

This also means that Coursera courses have naturally the highest "total enrollment" numbers, even though probably only a few percent of the students are actually taking the class.

I don't think this is the case will _All_ courses... Most of the CS courses allows you to access resources even after the course is completed.

Also if the course you want to enroll is not scheduled for near future, you can go to the course page and view the older version of the course.

I know I have had my eye on a few courses just to do the class.

I usually can find the course material online in a tar or zip file which has the home work and answers. Youtube usually has the videos. Still prefer to just open up the Coursera classes.

Coursera courses don't always close: I've been doing the compilers recently one and I've not come across any restrictions on access even though the last session was a year and a half ago.

Isn't it possible to just right-click the video and save a copy of it?

Last I looked at a Coursera course, there are explicit download buttons as well. It's not hard to save the course materials while the course is ongoing, the problem is that the materials are unavailable after the course finishes. You're not always going to know what will or will not be important in the future.

The Design Of Computer Programs on Udacity by Peter Norvig is pretty awesome. As someone who's only been seriously programming for a year or two , each problem set went like this :

Write the code here to do X :

I write a messy 4 liner after lots of thinking.

Professor Norvig comes along and does it in a simple functional one liner.

Mind blown.

He teaches good functional style program design one epiphany at a time.

  >> Professor Norvig comes along and does it in a simple functional one liner.
  >> Mind blown.
That is how I felt taking Martin Odersky's Scala course.

Also look at the sudoku solver code by peter norvig on his website. He explains it in a very simple way for a beginner and his code is short and simple.


It's truly the best course I've ever taken for newbies to seriously up their skills. Can't recommend enough.

Looks great. Will take a look at it when I learn Python.

I've completed a bunch of Coursera courses. Quality really varies. Even within the 9 course Data Science specialization [1] track some courses were rather poor while the rest were very good. I'm currently taking the #5 rated course [2]. It is excellent. But I'm only taking it because the Statisical Inference course in the Data Science specialization was so weak.

I would also recommend the Cryptography 1 course by Dan Boneh on Coursera [3]. Excellent if you are at all interested in the subject.

I always download the lecture videos, slides, quizzes, labs and exams because, as mentioned, many of the courses don't allow access once the class is completed.

You definitely have to have plenty of self discipline to complete MOOCs. And I don't have any delusions about a Coursera certificate being useful in landing a job; that's not what I'm after. I'm building the skills I want to apply to my own projects.

[1] https://www.coursera.org/specializations/jhudatascience [2] https://www.coursera.org/course/statistics [3] https://www.coursera.org/course/crypto

https://www.coursetalk.com has quite good reviews, especially on the more popular courses.

I am surprised that Khan Academy is not recognized in this report probably because the scope is limited to just those offer by an accredited university. I think calculus and chemistry have helped many first and second years of college students taking introductory courses including myself. I think KA is probably the most popular MOOC for all ages, in the most accessible way given it only requires a YouTube account, which millions have for over a decade. But that's just my opinion.

YouTube has a few really amazing courses available online.

For example, UCB has a channel with up-to-date content that otherwise not available on MOOC platform.

* https://www.youtube.com/watch?v=QMV45tHCYNI is a very good class on data structure

* https://www.youtube.com/watch?v=HyUK5RAJg1c and the rest has very good lectures on theory of computation

* https://www.youtube.com/watch?v=_G6_-ljgmXE also very good for algorithms. I find MIT's version to be too theory based for practitioners. Anyway, I still watch MIT's just to complement anything missing (no two speakers can teach the same topic equally)

Interesting but... I enrolled and completed a couple of the courses listed there. I'd be interested in seeing for each what percentage of enrolled students actually completed the whole course. Anecdotally the numbers are really really low (as in "around 5%").

I'm a huge MOOC fan, but after completing a couple of the very early Stanford ones I haven't completed any. Quite frankly I don't see completing a MOOC as the best use of time. I use MOOCs to either pick up new skills I need for work, brush up on stuff I haven't looked at since University or as 'entertainment', basically learning interesting stuff purely for fun. For non of these things actually completing the course to their satisfaction adds anything to their value.

I wish people would stop looking so much at completion scores and not judge the concept as a failure just because they are so low.

I would agree with and extend your remarks that getting a title of nobility from a brick and mortar uni gives you a vocational meal ticket, so you can motivate grinders by asking if they like money of if they like eating food, and then they'll grind away for letters on a transcript.

But in MOOC land I get nothing. I did the complete automata class mostly to see if there's much new since I did a brick and mortar version some decades ago, and the "reward" was not exactly fulfilling. I'm not expecting Ullman to fly out and shake my hand, or my phone to ring off the hook from recruiters seeing I jumped thru some hoops, but something more would have been nice? Actually, anything would have been nice?

Since the grind factor was aggressively burned out of me, I haven't completed a single course. I did watch the videos and did the assignments for computational neuroscience and one of the multi-class algo series and the functional programming and the famous AI class.

There's a lot of talk about how grinders don't grind therefore grinders suck or grinding sucks. However what really sucks is the playground presented to the grinders. Its not their fault they aren't being given something worth grinding.

By analogy most MMORPGs that tanked and died, tanked and died because the grind experience sucked. Its not the fault of the grinders that the businessmen couldn't or wouldn't present something as enjoyably grindable as WoW or Eve.

To some extent the whole market segment of MOOCs smells like a freshmen psych class lab experiment, "what happens to grind performance when you leave all the annoyance factors of grinds in place but remove all the rewards, with the hypothesis that almost every rat in the box stops grinding"

It's not clear what you're looking for exactly but, if anything, MOOCs (Coursera at any rate) have cut back on issuing completion certificates in the absence of paying a fee. Their problem, one of them anyway, is that certificates basically aren't worth anything as a certification mechanism in most cases so very few are willing to pay for them.

And, I expect, that if you start from the assumptions that 1.) MOOC certificates aren't worth anything as a real-world "vocational meal ticket" beyond the value of the underlying education and 2.) Companies offering MOOCs are not going to become unicorns and probably won't even return investors money-- you probably end up with something that isn't designed to mirror a university course.

I know that's the conventional top down narrative and I think you presented it well.

I'm proposing that as a bottom up narrative, they continue to include extensive grind barriers such as peculiar scheduling and pacing, someone else designing arbitrary barriers in terms of assignments with firm due dates, etc. Yet the reward has been cut out as per reasons in the top down narrative, etc. So from the bottom up perspective you're left with a stereotypical grind game where the reward has been removed, and from that theoretical model, the dropout rate makes sense. Then when people sign up and get something out of partial participation, it makes no sense to use a successful grind game metric measurement when its a failed grind game therefore the players aren't cooperating because they feel no need to do so. Because "MOOC as a grind game" has failed, a MOOC is a place to learn, not a place to jump thru hoops or not get a reward other than making someone else's completion metric result look good.

You can run a grind game without a vocational meal ticket reward model. WoW and Eve do just fine.

My suggestion would be to use advanced technology to abandon the grind game aspects, assuming higher completion rates are inherently valuable. Even if completion rates are not valuable, removal of pointless grind will result in better operating conditions for the students. Why must class schedules correspond vaguely to the northern hemisphere agricultural growing season? If you're not teaching project management, why are due dates so important? Why is there (typically) only one very fixed learning track if more exist or there is a "chose two from column B and three from column C" aspect inherent in the field itself? All of that would be very difficult to implement at Harvard in 1640, but due to technological advances would not be very challenging to implement today.

Maybe. I'd just say though that one person's "designing arbitrary barriers in terms of assignments with firm due dates, etc." is another's structured approach to learning. After all, there are plenty of learning opportunities if you want total flexibility.

Of course, as you suggest, there are potentially ways to gamify and provide incentives for learning that don't as closely mimic traditional course structures.

I completed a bunch of MOOC's. I've also worked on a bunch and not completed them while still learning. At some point, the learning took precedence and that's fine. I am profoundly grateful that there are free courses at the other end of the internet and feel lucky to have lived to seen the time in which it happened.

Yeah, not sure why completion matters.

All that matters is whether the user found some useful information in it, and just fully watching one video lecture should be enough for that metric.

Well, in a couple of cases (here is my current list of MOOCs I took part in, if anyone cares: http://pa-mar.net/Main/Study/Onlinecourses.html) in order to get completion you had to submit a end-of-course project which was very interesting in itself, and you also learned a lot because in order to get it "graded" you had to grade at least 5 other projects randomly selected among the other participants.

So you either saw different approaches to the same problem (e.g. Data Analysis) or everyone discussing a different problem (An Introduction to Operations Management). I am sure that I got a lot (in terms of learning) from that part of the course.

If everyone just picks and chooses some papers/videos and drops out before the end you don't get anyone to grade your work and you have no other work to grade (and possibly learn from).

Maybe at one point, considering our era, people reach full independence and can pick a book and a problem, with an ad-hoc study group and/or co-working space and just do it.

5% sounds pretty high to me, considering how low (nonexistent, really) the entry barrier is. Where does that figure come from, if I may ask?

I think I took it from an article online, and it matches my own experience (when you actually finish one of these there is often, but not necessarily always, a message from the course manager with some statistics about how many people turned in the final assignment or are anyway eligible for the certificate).

EDIT: Found a couple of articles:



Thanks for this. Incidentally, the second paper you link to is co-auhored, among others, by Daphne Koller, who teaches [this great course on probabilistic graphical models](https://www.coursera.org/course/pgm) and Andrew Ng, who teaches [the best-known intro MOOC in machine learning](https://www.coursera.org/learn/machine-learning/home/welcome).

Low barriers to entry would bump the completion rate down, as people are less invested in finishing the course. That's one of the reasons that college completion rates are usually >90%.

I did follow a few courses on coursera and I remember the professor saying that the number of students who successfully completed the first/previous iteration of the course was around ~5% or less, so I'd say you are right on the money.

I'll just throw in a plug here for Prof. Abu-Mostafa's Learning from Data course at Caltech, which is outstanding: http://work.caltech.edu/telecourse.html

Unlike Andrew Ng's Coursera Machine Learning class, it is a real, unadulterated Caltech class, and exactly as challenging as that implies. It delves much more deeply into the mathematics behind ML, and the homework assignments are quite time consuming.

I took it as part of an interactive session through EdX, and the professor himself was extremely active on the forums: responding to student questions, clarifying lecture points, and giving homework suggestions--seemingly at all hours of the day and night.

I did #27 "Introduction to Mathematical Thinking" during Feb-April this year. I enjoyed it a lot and found it challenging without being impossible to complete. Finished with a distinction and a feeling of great satisfaction.

I'm now doing Coursera's Interaction Design specialisation [1], which is proving to be very informative and a lot of fun.

If you're considering doing a MOOC then I'd definitely recommend it. Choose a free, short-ish course to start with, make the commitment, and dive in.

[1] https://www.coursera.org/specializations/interaction-design

That looks like a great series of courses. Unfortunately I am too busy with real school right now, but hopefully they offer them again.

I found it very interesting that the second most popular Mooc of all time is a philosophy course.

Although personally I like better the philosophy introduction offered by MIT:


I am also looking forward to this one https://www.edx.org/course/philosophy-minds-machines-mitx-24...

> An introduction to philosophy of mind, exploring consciousness, reality, AI, and more. The most in-depth philosophy course available online.

> What you will learn

The basics of argumentation

Some central arguments for and against the view that a sufficiently powerful computer can think (AI)

The main theories of mental states and their relations to physical states

Some central arguments for and against the view that the world is not as we perceive it to be

What the "hard problem of consciousness" is

I'm a big fan of uDacity, which only appears once in that list.

For a novice, their 'Intro to Computer Science' course is fantastic, as is the follow-on 'Web Development' course, led by Steve Huffman.

I've been doing Coursera courses for the past few years in an on/off fashion. I'm pleasantly

surprised by the courses that are more popular in this list.

I think the pattern is that the foundational or introductory courses are popular as they have a

larger audience. But it doesn't comment on the quality of the course. An interesting data point

is the social media "Share" widget that appears on the right column [0].

[0] https://www.coursera.org/course/rprog

I wonder how many of these were actually completed...

Enrolment would be an empty number to me, completion would be the mark of quality.

It's lower, but still orders of magnitude greater than complete the equivalent bricks and mortar course. If 1% complete the Android course that's 50 years worth of two classes of 30 students a year. At 10% it would take Adam Porter 500 years.

The thing about MOOC's is that scale is staggering.

It's insufficient to measure Popularity as the total number of registrants. People browse through course offerings, register, and may never return.

Rating by the total number of students that completed a course would be even more revealing. Add a student retention rate, too. Etc etc.

I'm not sure whether the number of people enrolled in a course is a very good measure for the impact of a course. I agree with some other commenters that the number of people that completed a course isn't necessarily a good measure either. Some measure that includes the percentage of videos watched, exercises attempted, links followed, etc. would probably better represent the actual impact of a course.

Some considerations: all courses have lots of people that enroll to subsequently discover the course is hard, boring or otherwise not what they expected. I expect this number varies strongly per course. On the other hand, not attempting/passing exercises doesn't mean someone hasn't invested time in watching all lectures.

It's heartening there are so many math focused courses. With the Calculus course at https://www.coursera.org/learn/calculus1/ - I'm going over differential calculus right now, but I'm getting somewhat stuck on the Chain Rule, and in particular working out how to use it to differentiate 2^x.

The courses I've found so far haven't been all that helpful, even Khan academy is confusing me somewhat. Does this course explain things better?

Since 2 = e ^ ln 2, use the identity:

2^x = (e^(ln 2))^x = e ^ (x * ln 2)

Well, that's the specific example, but there exist a class of educational blocks such that the learner doesn't know where they're stuck, because if they could define it, the act of correctly defining the stuck step would inherently unstick it.

So MOOC strategy works for some areas that don't have sticking points like that, but not so well where five minutes with a Socratic strategy tutor could save a lot of video re-watching time. Arguably, better video would point out likely sticking points. However, advanced math instructors got their position not by getting education degrees which theoretically vocationally train people to present like that, but got their position by being very successful when getting a PHD in an obscure and advanced part of the field. So if you're lucky, your higher math instructor might be a talented educator in addition to being a talented mathematician... but probably not, unfortunately.

It would be interesting as a thought experiment, or maybe as a startup, to see a good (emphasis on good) K-12 math instructor with a strong background in educational teaching skills try to teach diffeqs or higher math in general. I suspect they'd be extremely good at it, although probably very slow.

A lot of those questions are taken care on the forums. People tend to stumble on the same sticking points. But it part of learning to learn in this new medium.

Yeah, I get that bit, it's the next part I always get confused by. I'm sure I'll get it in the end. :-)

If you haven't yet, watch the Khan Academy videos on the chain rule. Then use the related practice questions until you've mastered them. That's how I got the concepts drilled into my head for almost all of Calculus I and II.

FWIW, I found the proof here:


I'd actually tried to understand it earlier, but I got stuck at the point it says:

  We now use these equations to rewrite f (g(x + h)). In
  particular, use the first equation to obtain
    f(g(x + h)) = f(g(x) + [g′(x) + v]h),

  and use the second equation applied to the right-hand-side
  with k = [g′(x) + v]h and y = g(x). 
It was how they got k that tricked me. But then I realise that actually the bit I was missing was that y=g(x), and so as:

  f(g(x + h)) = f(g(x) + [g′(x) + v]h)
You can actually make this:

  f(y + k) = f(y + [g'(x) + v]h)
Thus k = (g'(x) + v)h

That's kind of where I was getting confused :( I'm having a read of Better Explained at the moment:


Here is the list without 50 pagedowns:

  1. Programming Mobile Applications for Android Handheld Systems – Part 1 / University of Maryland
  2. Introduction to Philosophy / University of Edinburgh
  3. Inspiring Leadership through Emotional Intelligence / Case Western Reserve
  4. Introduction to Computer Science / Harvard University
  5. Data Analysis and Statistical Inference / Duke University
  6. Gamification / University of Pennsylvania / Wharton
  7. Social Psychology / Wesleyan University
  8. Circuits and Electronics / MIT
  9. Think Again: How to Reason and Argue / Duke University
  10. Creativity, Innovation and Change / Penn State
  11. A Beginner’s Guide to Irrational Behavior / Duke University
  12. Learn to Program: The Fundamentals / University of Toronto
  13. Game Theory / Stanford University, University of British Columbia
  14. Greek and Roman Mythology / University of Pennsylvania
  15. Startup Engineering / Stanford University
  16. Computational Investing, Part I / Georgia Institute of Technology
  17. Financial Markets / Yale University
  18. Introduction to Artificial Intelligence / Stanford University
  19. Introduction to Computer Science and Programming / MIT
  20. Introduction to Financial Accounting / University of Pennsylvania / Wharton
  21. Modern & Contemporary American Poetry / University of Pennsylvania
  22. Machine Learning / Stanford University
  23. Data Analysis / Johns Hopkins Bloomberg School
  24. Introduction to Computer Science and Programming Using Python / MIT
  25. Science and Cooking: From Haute Cuisine to Soft Matter Science / Harvard University
  26. Introduction to Philosophy: God, Knowledge, and Consciousness / MIT
  27. Introduction to Operations Management / University of Pennsylvania / Wharton
  28. Introduction to Mathematical Thinking / Stanford University
  29. Justice / Harvard University
  30. A History of the World Since 1300 / Princeton University
  31. Creative Programming for Digital Media & Mobile Apps / University of London/ Goldsmiths
  32. Neural Networks for Machine Learning / University of Toronto
  33. Learn to Program – Crafting Quality Code / University of Toronto
  34. Critical Thinking in Global Challenges / The University of Edinburgh
  35. Statistics – Making Sense of Data / University of Toronto
  36. Introduction to Biology – The Secret of Life / MIT
  37. Drugs and the Brain / Caltech
  38. Introduction to Databases / Stanford University
  39. The Ancient Greek Hero / Harvard University
  40. Social Network Analysis / University of Michigan
  41. Health in Numbers: Quantitative Methods in Clinical & Public Health Research / Harvard University
  42. Introduction to Astronomy / Duke University
  43. Human Health and Global Environmental Change / Harvard University
  44. Software Defined Networking / Princeton University
  45. Introduction to Statistics: Descriptive Statistics / UC Berkeley
  46. Computing for Data Analysis / Johns Hopkins Bloomberg School of Public Health
  47. Functional Programming Principles in Scala / Ecole Polytechnique Federale de Lausanne
  48. The Camera Never Lies / University of London/ Royal Holloway
  49. Calculus One / Ohio State University
  50. Maps and the Geospatial Revolution / Penn State

I typically hate ranking lists like this, since I believe student completion rates are more important than enrollment. However, I did find some joy in finding a new MOOC I could add to the backlog of courses I need to take.

I'm not sure why you'd worry about completion rates. When I've taken online courses I've completed all the lectures and their associated exercises. But if I'm taking a course it means I have some end in mind for the skills it teaches me and that makes any final project or exam superfluous.

I have to disagree: for me the final project was important to immediately put to the test the things I had learned. And - as I explained above - if the course has peer-grading (almost a given if it has a final project, because you cannot automate the evaluation of a project) you learn even more by evaluating the ways other approached their solution.

I'm a little surprised that there is no mention of iTunes University, or maybe that doesn't count as a MOOC. I know there are some very popular courses on there that are easy to access.

iTunes university is really quite awesome. Only you need an Apple product to access it.

iTunes runs on Windows

There are also bunch of third party apps for Android and Linux that let you download iTunes U lectures.

I didn't know it existed, I'll have to check it out.

Surprise that a philosophy class ranked at number two. Other then that the list mostly looks like what you'd expect.

I'm surprised, no Jeff Ullman Automata?

It would be nice if the prerequisites for a course were listed more clearly.

I don't like education and universities and all that stuff, I would like them to release the videos and the text in the open, that I can learn my way, not in a supervised fashion, behind a login wall.

there's a MOOC aggregator to help keep on top of course offerings and it tracks programs globally: www.class-central.com

Applications are open for YC Summer 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact