Sifting Through Data With A Neural Net

IMAGINE COMPRESSING TV pictures and high-fidelity music so tightly that they could travel simultaneously over ordinary phone lines. Robert Hecht-Nielsen, chief scientist at HNC Software Inc. in San Diego, thinks he's on the track of just such a breakthrough--using so-called neural networks. These are brainlike circuits that learn from experience and don't need conventional programming.

To continue reading this article you must be a Bloomberg Professional Service Subscriber.