No, this is not the repetitive chant of a zombie… although it could be. However, what I’m referring to, is Google’s latest acquisition in the area of great minds. I’m referring to the DNNresearch company led by Prof. Geoffrey Hinton, who’s research in the field of convolutional neural networks has attracted quite some attention in the area of Machine Learning. And this is just the latest of a continuing trend in Google’s plan to attract the greatest researchers to push the current boundaries. I wonder what “online” will become in 20 years.
Archive for the ‘Machine Learning and AI’ Category
Emotions are a major part of what makes us human. So why is this part almost neglected in this informational era? Here‘s an article that highlights the new trends in IT, suggesting that emotions have to and will play an increasingly important role in how we interact with machines.
Do I need to say more? Prof. Jeoffrey Hinton is one of the best known researchers in the field of Machine Learning and AI. Enjoy the clip.
As they say, this tool seems ideal for tests in terms of architecture and management of emergencies.
A short documentary, called “In its image”, about expert systems, genetic algorithms, neural networks, the perceptron, and much more.
The Economist writes about deep learning and convolutional neural networks in Computer Vision.
How does our brain work…? New theories.
Can we reverse engineer the human brain? Some say yes, and even so early as in 10 years from now. And there’s important research being done in that direction too, at EPFL for example. In the mean time, here are the main ideas:
Here are a couple of extra videos from the Blue Brain project, visualizing the neurons and their structure in a way that gives you (well, at least me right now) a lot of new information and insight.
There are some, more or less, new ideas spreading throw the world of computer science, in particularly fields like A.I., Machine Learning or Visualization, fields that are in some way connected to mimicking, analyzing or converting the human brain in a virtual system.
I decided to roll two of these ideas out here today, in very simple terms, that are a good starting point of a brainstorming session on these issues. Because the way/angle in which you look at something will define how you see it, and if you look at it correctly there’s a larger chance that you will discover the features that will power you forward in you endeavor.
Jeff Hawkins from the University of Berkeley, California, says it very clearly: “Intelligence is defined by prediction”.
On the other hand, the brain can only do so much in abstract terms. So, as you may have noticed, it uses other features a lot as a support for the day by day tasks – measures like visualizing. The tremendously cool part is, that if you think about it, you can consider visualization as outsourcing the main load of the computation, to an area of the brain capabilities that is able to cope with much larger sets of data, as in visual representations.
An important part of understanding artificial intelligence or machine learning algorithms and development ideas is given by the comprehension of our selves, of your brains.
So if most of the devices we intend to create should learn as we do, than looking deeply into the way out brains work should be a vital first step for the process, for any AI or Machine Learning researcher.
To better understand what is happening and to grasp the state of the art in neuroscience, I recommend following these interesting presentations about the human brain and the way it processes data:
Very interesting ideas!
By the way, if you haven’t heard about it yet, you should definitely visit www.TED.com, a website with many such interesting scientific presentations.