|Learning From Data - Live From Caltech|
|Written by Sue Gee|
|Friday, 16 March 2012|
Do you fancy joining in a course taught by Caltech Professor Yaser Abu-Mostafa? If so you can sign up as a remote student to join in lectures twice per week. The catch is that the lectures are live and you have to "be there" even if it is the middle of the night for you.
The latest entry on Class Central's list of free online computer-science related courses is one with a difference.
The lectures for Learning With Data, an introductory Machine Learning course, are to be delivered as a live broadcast at the same times on Tuesdays and Thursdays throughout April and May. The only lectures that will be available recorded will be the two from the first week, to allow for late-joining students.
The course consists of 18 lectures, each 1 hour long followed by 25 minutes of Q and A. Students in the live broadcast audience will be able to text questions using a chat box next to the stream. There will be online homework with multiple choice questions and scoreboard ranking of participants.
Billed as a "real Caltech course" and "not a watered-down version", it sets out to balance theory and practice, and cover the mathematical as well as the heuristic aspects of machine learning. The prerequisites are: basic probability, matrices, and calculus.
The complete schedule of lectures to be delivered by Professor Yaser Abu-Mostafa has already been posted, with the starting times for different regions, and registration opens on March 26th.
This is a different model from the Coursera Machine Leaning Course from Stanford's Professor Andrew Ng, which has video lessons split into short segments interspersed with quizzes and has programming assignments and online exams as well as homework. No start date has yet been announced for its next presentation.
As our recent article on free online courses demonstrated there is currently no shortage of opportunities to learn computer science topics at a variety of levels.
Given the new modes of delivering courses online it is something of a mystery why time enters the equation at all. Why can't I just sign up for an online course and do it when I want to? The lectures are recorded and the assignments are marked automatically so why does there have to fixed start date?
On the other hand you could argue that live lectures are superior to canned talks. Perhaps there are just two viable models in the long term fully recorded courses that you can take anytime and the real live experience of being in a lecture.
If you are interested in machine learning now is your chance to find out.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Friday, 16 March 2012 )|