Synchronized brain waves enable rapid learning | MIT News Office

Synchronized brain waves enable rapid learning | MIT News Office:  Brain waves known as “beta bands,” produced independently by the prefrontal cortex and the striatum, began to synchronize with each other. This suggests that a communication circuit is forming between the two regions, Miller says.

“There is some unknown mechanism that allows these resonance patterns to form, and these circuits start humming together,” he says. “That humming may then foster subsequent long-term plasticity changes in the brain, so real anatomical circuits can form. But the first thing that happens is they start humming together..."

Previous studies have shown that during cognitively demanding tasks, there is increased synchrony between the frontal cortex and visual cortex, but Miller’s lab is the first to show specific patterns of synchrony linked to specific thoughts.

Robot elephant trunk learns motor skills like a baby - tech - 13 March 2014 - New Scientist

Robot elephant trunk learns motor skills like a baby - tech - 13 March 2014 - New Scientist: The design showed that a trunk formed of 3D-printed segments can be controlled by an array of pneumatic artificial muscles...

They used a process called "goal babbling"... the robot remembers what happens to the trunk's position when tiny changes are made to the pressure in the thin pneumatic tubes feeding the artificial muscles. This creates a map that relates the trunk's precise position to the pressures in each tube.

The trunk can now be manually forced into a series of positions and learn to adopt them on command...

Using large-scale brain simulations for machine learning and AI

Using large-scale brain simulations for machine learning and AI: Neural networks are very computationally costly, so to date, most networks used in machine learning have used only 1 to 10 million connections. “But we suspected that by training much larger networks, we might achieve significantly better accuracy,” said the Google team.

“So we developed a distributed computing infrastructure for training large-scale neural networks. Then, we took an artificial neural network and spread the computation across 16,000 of our CPU cores (in our data centers), and trained models with more than 1 billion connections.”

“We then ran experiments that asked, informally: If we think of our neural network as simulating a very small-scale ‘newborn brain,’ and show it YouTube video for a week, what will it learn? Our hypothesis was that it would learn to recognize common objects in those videos.

“Indeed, to our amusement, one of our artificial neurons learned to respond strongly to pictures of… cats. Remember that this network had never been told what a cat was, nor was it given even a single image labeled as a cat.

Music of the brain: each synapse has its own natural rhythm | KurzweilAI

Music of the brain: each synapse has its own natural rhythm | KurzweilAI: Contrary to what was previously assumed, Mehta and Kumar found that stimulating the neurons at the highest frequencies was not the best way to increase synaptic strength. “To our surprise, we found that beyond the optimal frequency, synaptic strengthening actually declined as the frequencies got higher.”

The knowledge that a synapse has a preferred frequency for maximal learning led the researchers to compare optimal frequencies based on the location of the synapse on a neuron...

The optimal frequency for inducing synaptic learning changed depending on where the synapse was located. The farther the synapse was from the neuron’s cell body, the higher its optimal frequency.

“Incredibly, when it comes to learning, the neuron behaves like a giant antenna, with different branches of dendrites tuned to different frequencies for maximal learning,” Mehta said.

Video: A Robot That Can Figure Out New Tasks Based On the Ones It Knows | Popular Science

Video: A Robot That Can Figure Out New Tasks Based On the Ones It Knows | Popular Science: To borrow an example from the video below, the robot can fill up a glass of water from a bottle via pre-programmed instructions. But if halfway through the task its overseer asks it to chill the water, the robot will actually stop and think about the next steps.

Figuring that it can’t grab an ice cube until it empties one of its two hands, it then reasons that the water bottle is more expendable than the glass of water and sets the bottle down. It then grabs the ice and drops the cube into the glass. Task completed, no extra programming necessary.

Tapping Quantum Effects for Software that Learns - Technology Review

Tapping Quantum Effects for Software that Learns - Technology Review: D-Wave's machine is intended to do one thing better than a conventional computer: finding approximate answers to problems that can only be truly solved by exhaustively trying every possible solution. D-Wave runs a single algorithm, dubbed quantum annealing, which is hard-wired into the machine's physical design, says Geordie Rose, D-Wave's founder and CTO. Data sent to the chip is translated into qubit values and settings for the couplers that connect them. After that, the interlinked qubits go through a series of quantum mechanical changes that result in the solution emerging. "You stuff the problem into the hardware and it acts as a physical proxy for what you're trying to solve," says Rose. "All physical systems want to sink to the lowest energy level, with the most entropy," he explains, "and ours sinks to a state that represents the solution."

Robots learning from experience (w/ Video)

Robots learning from experience (w/ Video): "An important development from Xpero is the robot’s ability to build its knowledge base. “It makes no distinction between previous knowledge and learnt knowledge,” explains Kahl. “That it can re-use knowledge is very important. Without that there would be no incremental learning.”
In award-winning demonstrations, robots with the Xpero cognitive system on board have moved about, pushed and placed objects, learning all the time about their environment. In an exciting recent development the robot has started to use objects as tools. It has used one object to move or manipulate another object that it cannot reach directly."