Professor
John Hopfield
Email:hopfield@Princeton.EDU
Web:
Hopfield and his collaborators are working to understand how
the neural circuits of the brain perform complex calculations.
The human brain is astonishingly effective at processing information
in many inexact problems computers find difficult, problems such
as recognizing a particular person in a crowded scene or understanding
natural language in a noisy environment. Understanding how the
brain carries out such calculations is a major challenge to neuroscience.
Through learning how a brain solves such problems, it may be possible
to build faster and more powerful manmade systems.
In 1982 he discovered a network metaphor or for how the brain
works, describing a computational process in terms of a dynamical
attractor. More recently he has been pursuing how to use action
potentials in computation. 100 billion of these are generated
each second in the human brain. Hopfield is trying to understand
how the times at which these occur can be used in the computational
process of the brain. This research is also linked to describing
how nerve cells process sensory information to turn it into perceptions,
such as odors or sounds.
Hopfield came to computational science from a background in physics.
After earning a Ph.D. from Cornell University in 1958, he next
took a position in the Bell Laboratories theory group. He segued
from physics to biology while a physics professor at Princeton,
finding useful conceptual relationships between semiconductors
and solids, which he had been studying as a condensed matter physicist,
and the properties of protein molecules. As he became more involved
with information processes in biology, he saw a need for formal
mathematics and theory in neurobiology, and began his current
line of research.
|