Five new Computer Science courses start in the next two weeks. Other free online classes in Computer Science and Electronics have already started and some are still open to latecomers.
The phenomenon of the free MOOC Massive Open Online Course in which the participants are distributed across the globe and the course materials are delivered via the internet isn't new. The Khan Academy has been offering free courses for over five years and from its beginnings teaching math now has a collection of more than 3,000 video lectures on YouTube in a long list of subjects.
Andrew Ng's Stanford class in Machine Learning had also been delivered via You Tube for some years before last autumn's experiment in which it added interactive features which meant that students could submit homework, take exams and earn a certificate.
This is an idea that has taken hold in Computer Science, perhaps because it is a subject where you can set realistic homework assignments that can be marked by a computer.
The class that attracted a record number of students for a MOOC was Sebastian Thrun and Peter Norvig's Introduction to Artificial Intelligence. Even if you discount the figure of 160,000 sign ups on the grounds that many people enrolled multiple times and a high proportion dropped out, the important statistic is that over 23,000 students from 190 countries graduated a demanding course that was indeed at "advanced undergraduate" level.
The experience of the AI course, and of the Machine Learning and Database courses that ran at the same time and also attracted huge numbers of enrollments (104,000 and 92,000, respectively) and large numbers of graduates (13,000 and 7,000) proved that there is a demand for education at this level - particularly when it is taught be established academics and is free of charge.
Currently there are three new initiatives to fill this demand, Udacity, Coursera and MITx.
Sebastian Thrun was so impacted by the AI course that he co-founded Udacity, an initiative backed by Charles River Ventures. Two computer courses from Udacity and already underway: the introductory level CS101: Building a Search Engine which teaches Python to non-programmers and CS 373: Programming a Robotic Car. As the 3 in the course code indicates, this is an "advanced course" which demands not only a facility with math and skill in Python programming, but also probability theory at the level achieved in last year's AI courses. Both these courses started on February 20th but, after recent publicity in the New York Times, are remaining open for new enrollments. Late joining students will have their final result based solely on the exam at the end of the course.
Four new Udacity Courses, starting on 16th April have now been announced. Three are at an intermediate level:
and CS212 - The Design of Computer Programs to be taught by Peter Norvig and for which no video introduction is yet available. The fourth course is at level 3 and is CS387 - Applied Cryptography being taught by David Evans, the tutor for the current CS101. Typically Udacity courses have six weeks of teaching with a seventh week for the exam.
No dates and course lengths are yet available for CS 101, Introduction to Machine Learning, Human-Computer Interaction, Computer Security and the courses in Entrepreneurship and other subjects that were originally announced last November but they are all "upcoming".
The first course from MITx is now coming to the end of its first week. 6.002x - Circuits and Electronics involves video lectures, interactive labs and homework and at 14 weeks is double the length of the he Coursera ones that range from 5 to 10 weeks. This too is an undergraduate level course and the difference between this initial beta presentation is that there is no charge for the certificate of accomplishment awarded for successful completion.
The first homework for this course is due in on March 16th so hurry to join in if you want an online grounding in this alternative computing science.
Amazon has lots and lots of warehouses, sorry fulfilment centers, all over the world and they employ lots of humans to find the stuff you buy. The Amazon Picking Challenge is about getting robots to d [ ... ]
We have recently seen how much neural networks can learn, but the latest example is very strange. Google has succeeded in teaching a neural network to be a computer simply by showing it programs and w [ ... ]