Neural networks modeled on the human brain have advanced over the past 20 years and are now making their way into mainstream computing, due to improvements in both hardware and software.
Although computers are still mostly incapable of independent thought, they can now process large amounts of data to reach basic conclusions without human assistance. Micron and IBM are developing hardware that can create more advanced neural networks, while software developments also are bringing neural networks to real-world settings. Google, for example, has used neural network algorithms to improve its Google Voice speech recognition application.
In neural networking, unlike traditional computing, the computer is primarily responsible for solving a specific problem on its own, notes Rochester Institute of Technology professor Leon Reznik.
Advances in silicon have improved neural networking by offering the requisite density to run large clusters of nodes even on a single slice of silicon. Although neural networks are unlikely to replace standard central-processing units (CPUs), they might take on tasks that are too difficult for CPUs alone to handle. "Instead of bringing sensory data to computation, we are bringing computation to sensors," says IBM's Dharmendra Modha. "This is not trying to replace computers, but it is a complementary paradigm to further enhance civilization's capability for automation."
From IDG News Service
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA