Introduction to Adaptive
and Neural Networks
What is an Adaptive
Network?
- a network of simple, interconnected processors that "learns from"
and "adapts to" changing conditions
- can be simulated on a general-purpose computer such as a PC or incorporated
in high-speed, low-cost, special-purpose electronic "chips"
What is a Neural
Network?
- The "real" Neural Networks are those biological networks invented
by nature to sense, learn from, and interpret the world, e.g., the brain,
the visual cortex, and the sensory-motor systems of living beings
- An "Artificial Neural Network" is a form of Adaptive Network that
is motivated by the architecture of biological networks; often the term Neural
Network means "Artificial Neural Network"; also, the distinction
between Adaptive and Neural Networks is generally blurred
When do you need
an Adaptive Network?
- to perform tasks that are difficult for other technical approaches
- to perform tasks that require "learning", "judgement"
or "intelligence"
- to infer or learn from "data-rich", "understanding-poor"
situations
- to perform tasks that require human attention, but are repetitious or hazardous
- to predict and control chaotic dynamical systems
Return to ANSR Home Page or Site
Contents -- or -- Contact ANSR.