Not sure if people remember but Stanford in collaboration with Sebastian Thrun offered one of the first online educational course in Machine Learning. This piqued my interest in ML and I enrolled & completed the course from one of the most rural parts of India (I had to create an extension antenna for my 3g phone to get better reception).
Udacity was born due to the popularity of this course.
Can't thank Sebastian and Stanford enough for this free course.
That's a [edit:perfectly fine] example to choose for a Stanford MOOC, since Coursera was started based on the success of Andrew Ng's machine learning course.
edit: People below (and above) are more knowledgeable than I am.
Actually, Thrun's AI course, Ng's ML course, and Widom's Databases course were announced and ran pretty much simultaneously. They were all successful, but only Thrun decided to split off because of it.
That course is a very good basic machine learning course. It has many of the stepping stones to complex machine learning problems, no matter which specific field.
So yes, it is a good intro to machine learning. There are new advances coming up all the time. But you will definitely need to know most of the topics in that course. That is, if you want to properly understand most of the latest techniques.
I'm curious why there hasn't been a push towards more efficient compiled languages like golang?
I don't work with machine learning and AI, but I used to do a lot of server-side programming with dynamic language. Switching to golang has been great. I'm far more productive with it and the CPU and memory savings have been great. Isn't ML the kind of domain where CPU cycles and memory matter?
No - AI Class (Norvig/Thrun) and ML Class (Ng) both ran at the same time, in the Fall of 2011. I took both courses, but only completed the ML Class (I had to drop the AI Class due to personal issues at the time).
I'm not sure which MOOC company started first (in 2012), though - I do think it may have been Udacity. I know that when it did start, it wasn't able to offer the AI Class because of {reasons} (I think maybe licensing of course content or something) - and so instead it offered the CS373 course (at the time titled "How to Build Your Own Self Driving Vehicle" or something like that). By that time (Spring 2012), I had moved on past my earlier problems, and jumped at that course - I took it and completed it as well.
Coursera, meanwhile, was able to offer the original ML Class as a premier course (maybe Ng had different rights to the content, or maybe there was issues on the AI Class that had to do with the dual instructor partnership of the class - I'm not sure what really happened there).
After about a year (IIRC?), Udacity was able to finally offer the original AI Class as part of their courses (I think now renamed "Introduction to Artificial Intelligence").
Today, I'm taking the Udacity Nanodegree course, as I've noted before here.
I think this offering from Stanford is interesting, but it doesn't seem like it is currently "available" to enroll, because the two required courses seem closed or something? Maybe they're taking enrollment for a future starting date. That said, while the tuition isn't outside of something I could do, I currently think the best use of my money and time - after I complete the Udacity thing - would be to pursue getting a BA on CompSci or something of that nature, then pursuing other paths.
Should've done it a long time ago when I was younger - but I was dumb.
Let's not do anything drastic. If you've finished those courses and have decent experience a bachelor's degree will get you little to no practical value. If there are holes in your knowledge, by all means fill them, but a degree of little value is not a good investment in this line of work.
You are better off filling holes in your knowledge and build something awesome. That's it. You want an interview & offer at Google, facebook, Microsoft, it's pretty easy with a solid portfolio and a few months study or refresh on algorithms.
Thrun & Norvig did Intro to AI, Ng did Machine Learning, Koller did PGM (another course related to AI/ML) and Widom did DB/SQL, which is how it all started. Thrun then started Udacity, Ng & Koller started Coursera.
Did that course actually. Sebastian Thrun and Peter Norvig were the instructors. Still have the certificate from the course(back then they used to provide signed certificates from the instructors for free).
Udacity was born due to the popularity of this course.
Can't thank Sebastian and Stanford enough for this free course.