Magnets can help AI get closer to the efficiency of the human brain

Purdue University researchers have developed a process to use magnetics with brain-like networks to program and teach devices to better generalize about different objects.

Credit: Kaushik Roy/Purdue University

Computers and artificial intelligence continue to usher in major changes in the way people shop. It is relatively easy to train a robot's brain to create a shopping list, but what about ensuring that the robotic shopper can easily tell the difference between the thousands of products in the store?

Purdue University researchers and experts in brain-inspired computing think part of the answer may be found in magnets. The researchers have developed a process to use magnetics with brain-like networks to program and teach devices such as personal robots, self-driving cars and drones to better generalize about different objects.

"Our stochastic neural networks try to mimic certain activities of the human brain and compute through a connection of neurons and synapses," said Kaushik Roy, Purdue's Edward G. Tiedemann Jr. Distinguished Professor of Electrical and Computer Engineering. "This allows the computer brain to not only store information but also to generalize well about objects and then make inferences to perform better at distinguishing between objects."

Roy presented the technology during the annual German Physical Sciences Conference earlier this month in Germany. The work also appeared in the Frontiers in Neuroscience.

The switching dynamics of a nano-magnet are similar to the electrical dynamics of neurons. Magnetic tunnel junction devices show switching behavior, which is stochastic in nature.

The stochastic switching behavior is representative of a sigmoid switching behavior of a neuron. Such magnetic tunnel junctions can be also used to store synaptic weights.

The Purdue group proposed a new stochastic training algorithm for synapses using spike timing dependent plasticity (STDP), termed Stochastic-STDP, which has been experimentally observed in the rat's hippocampus. The inherent stochastic behavior of the magnet was used to switch the magnetization states stochastically based on the proposed algorithm for learning different object representations.

The trained synaptic weights, encoded deterministically in the magnetization state of the nano-magnets, are then used during inference. Advantageously, use of high-energy barrier magnets (30-40KT where K is the Boltzmann constant and T is the operating temperature) not only allows compact stochastic primitives, but also enables the same device to be used as a stable memory element meeting the data retention requirement. However, the barrier height of the nano-magnets used to perform sigmoid-like neuronal computations can be lowered to 20KT for higher energy efficiency.

"The big advantage with the magnet technology we have developed is that it is very energy-efficient," said Roy, who leads Purdue's Center for Brain-inspired Computing Enabling Autonomous Intelligence. "We have created a simpler network that represents the neurons and synapses while compressing the amount of memory and energy needed to perform functions similar to brain computations."

Roy said the brain-like networks have other uses in solving difficult problems as well, including combinatorial optimization problems such as the traveling salesman problem and graph coloring. The proposed stochastic devices can act as "natural annealer," helping the algorithms move out of local minimas.

Their work aligns with Purdue's Giant Leaps celebration, acknowledging the university's global advancements in artificial intelligence as part of Purdue's 150th anniversary. It is one of the four themes of the yearlong celebration's Ideas Festival, designed to showcase Purdue as an intellectual center solving real-world issues.

Roy has worked with the Purdue Research Foundation Office of Technology Commercialization on patented technologies that are providing the basis for some of the research at C-BRIC. They are looking for partners to license the technology.

Team Work - Meaning and Tips for better Team Work

A single brain is not always capable of making key decisions on its own. To come up with an efficient solution, an individual requires the help and advice of others. A team is established when individuals get together on a common platform with the common goal of completing a task. To guarantee optimum compatibility, team members should ideally come from similar backgrounds and have a single aim. To provide their best, the team members must complement each other and function as a single unit in tight cooperation. "There is no I in Team Work," as the saying goes, and each member must put the needs of his team first. Personal interests must take a second seat. Any team's performance is directly proportionate to the relationship between its members and their combined efforts. What is the definition of teamwork? Teamwork is defined as the sum of each team member's efforts toward the fulfilment of the team's goal. In other words, any team's backbone is its ability t

Scientists discover a new theory / The fundamental property of light – 150 years after Maxwell

Light plays a vital role in our everyday lives and technologies based on light are all around us. So we might expect that our understanding of light is pretty settled. But scientists have just uncovered a new fundamental property of light that gives new insight into the 150-year-old classical theory of electromagnetism and which could lead to applications manipulating light at the nanoscale. It is unusual for a pure-theory physics paper to make it into the journal Science. So when one does, it’s worth a closer look. In the new study, researchers bring together one of physics' most venerable set of equations – those of James Clerk’s Maxwell’s famous theory of light – with one of the hot topics in modern solid-state physics: the quantum spin Hall effect and topological insulators . To understand what the fuss is about, let’s first consider the behaviour of electrons in the quantum spin Hall effect. Electrons possess an intrinsic spin as if they were tiny spinning-tops,

19 Types Of Content Writing Services For Your Business

It’s hard to know which type of content writing service is the best for your business. There are so many different types of content writing services out there that it’s easy to get confused. You end up wondering if you’re choosing the right one for you. In this post, we’ll get rid of this confusion, once and for all. I’m going to list out the different kinds of writing services you could use. By the end of this article, you’ll know whether you need a copywriter, a content writer, or a social media marketer and how they can help you achieve your business goals. This post is also useful for writers who want to hone their writing skills in a specific area. Let’s dive in and learn what types of content writing services exist and when you should use them. (Bonus – if you want to hire the top 1% of writers, go to the bottom to learn how). Types of Content Writing Services As we go through the list of content writing services, you will find that many of them overlap. That’s perfectly

The LimitlessX - Get Tech Science News around the Globe

Search This Blog