قالب وردپرس درنا توس
Home / Science / Israeli researchers develop a brain-inspired AI algorithm – HEALTH & SCIENCE

Israeli researchers develop a brain-inspired AI algorithm – HEALTH & SCIENCE



  Processing an event with multiple objects.

Processing an event with multiple objects. ,
(Credit: by courtesy)

Bar Ilan University researchers said they had developed a new type of ultra-fast artificial intelligence (AI) algorithm based on the dynamics of the human brain.

According to a report published on Friday in Scientific Reports the algorithm shows that the human brain is extremely fast and efficient despite its much slower computational speed compared to modern computers. As such, the scientists say: "The insights of the basic principles of our brain must once again be at the center of future artificial intelligence."

In particular, researchers headed by Prof. Ido Kanter of the Bar Ilan University Institute of Physics and Gonda Multidisciplinary Brain Research Center used advanced experiments on large-scale neuronal cultures and simulations to demonstrate a new algorithm that could be used Learning outperforms rates previously achieved with state-of-the-art learning algorithms.

"The current scientific and technological view is that neurobiology and machine learning are two distinct disciplines that proceed independently," said Kanter. "The lack of an expected mutual influence is puzzling."

Kanter explains that the number of neurons in a brain is less than the number of bits in a modern-size computer, and that the brain's computational speed is slower than that of the first computers invented more than 70 years ago. However, the learning rules of the brain are much more complicated than those of the current AI algorithms.

"While driving, one observes cars, pedestrian crossings and traffic signs and can easily identify their timing and relative positions," Kanter said. Biological Hardware [learning rules] was developed to process asynchronous inputs and refine their relative information.

This is in contrast to traditional AI algorithms, which rely on synchronous inputs and ignore the relative timing of various inputs that make up the same frame. [1

9659005] Kanter's Conclusion: The drawback of the brain's complicated learning scheme can be beneficial .

`;
document.getElementById ("linkPremium"). innerHTML = cont;
(Function (v, i) {
});



Source link