Friday, 22 December 2017

Memristors power quick-learning neural network

A new kind of neural network created with memristors can vividly enhance the efficiency of teaching machines to think like humans. The network, called a reservoir computing system, could predict words before they are said during conversation, and assist to predict future results based on the current. The research team that developed the reservoir computing system, led by Wei Lu, professor of electrical engineering and computer science at the University of Michigan, recently published their work in Nature Communications.

Reservoir computing systems, which enhance the typical neural network’s capacity and lower the required training time, have been developed before with bigger optical components. Nonetheless, the U-M group developed their system using memristors, which need less space and can be integrated more easily into existing silicon-based electronics.
Memristors are a special type of resistive device that can both perform logic and store data. This contrasts with typical computer systems, where processors process logic independent from memory modules. To train a neural network for a task, a neural network takes in a large set of questions and the answers to those questions. In this process of what’s called supervised learning, the connections between nodes are weighted more heavily or lightly to reduce error in getting the correct answer.
After training, a neural network can then be tested without knowing the answer. For example, a system can process a new photo and correctly identify a human face, because it has learned the features of human faces from other photos in its training set.

No comments:

Post a Comment