November 24, 2012
Big Strides With Deep Learning Computer Systems

Want to improve human learning systems? That's so 20th century. Machine learning is all the rage.

But what is new in recent months is the growing speed and accuracy of deep-learning programs, often called artificial neural networks or just “neural nets” for their resemblance to the neural connections in the brain.

“There has been a number of stunning new results with deep-learning methods,” said Yann LeCun, a computer scientist at New York University who did pioneering research in handwriting recognition at Bell Laboratories. “The kind of jump we are seeing in the accuracy of these systems is very rare indeed.”

One of the most promising points in the article was about a deep learning system that a university team put together quickly to win a contest on how to discover promising drug candidates. While computer automation is on course to automate many more kinds of jobs the ability to speed up biomedical research holds the most promise to deliver big benefits for us all.

Share |      Randall Parker, 2012 November 24 05:23 PM 

James Bowery said at November 25, 2012 2:36 PM:

The deep learning systems need to make some progress in pruning neurons so their generalization is better. That is to say, they seem prone to over-fitting. This boils down to understanding the relationship between Kolmogorov Complexity and intelligence, and applying it. The good news is that we might finally be starting to understand the neocortex's 8 (or so) layers. Too bad we've been invaded. Maybe the AIs will come along soon enough to start saying things that a 400 IQ child would say. What are they going to do? Call it a "neo-Nazi hater"?

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©