[ad_1]
ANI |
Updated: Apr 23, 2020, 3:48 PM IST
Washington DC. [USA], April 23 (ANI): The researchers reveal that they have successfully rebuilt the bridge between experimental neuroscience and advanced artificial intelligence learning algorithms. By conducting new types of experiments in neural cultures, the researchers were able to demonstrate a new accelerated learning mechanism inspired by the brain.
In an article published in the journal Scientific Reports, researchers say that when the mechanism was used in the artificial task of handwritten digit recognition, for example, its success rates substantially outperformed commonly used machine learning algorithms.
To reconstruct this bridge, the researchers set out to test two hypotheses: that the common assumption that learning in the brain is extremely slow could be wrong, and that the dynamics of the brain could include accelerated learning mechanisms. Surprisingly, both hypotheses were proved correct.
“It is believed that a learning step in our brain tends to last tens of minutes or even longer, while on a computer it takes a nanosecond, or a million times faster,” said study lead author Professor Ido Kanter, de Bar. -Department of Physics at the University of Illan and Multidisciplinary Center for Brain Research Gonda (Goldschmied). “Although the brain is extremely slow, its computational capabilities exceed, or are comparable to, typical cutting-edge artificial intelligence algorithms,” added Kanter, who was assisted in the research by Shira Sardi, Dr. Roni Vardi, Yuval Meir, Dr Amir Goldental, Shiri Hodassman and Yael Tugendfaft.
The team’s experiments indicated that adaptation in our brains accelerates significantly with training frequency. “Learning by observing the same image 10 times in a second is as effective as observing the same image 1,000 times in a month,” said Shira Sardi, the main contributor to this work. “Repeating the same image quickly improves adaptation in our brains to seconds rather than tens of minutes. Learning in our brains may be even faster, but beyond our current experimental limitations,” added Dr. Roni Vardi, another main contributor to the research. . Utilizing this newly discovered brain-inspired accelerated learning mechanism substantially outperforms commonly used machine learning algorithms such as handwritten digit recognition, especially when providing small data sets for training.
The reconstructed bridge from experimental neuroscience to machine learning is expected to advance artificial intelligence and especially ultrafast decision-making under limited training examples, similar to many circumstances of human decision-making, as well as robotic control and optimization of network. (AND ME)
[ad_2]
