Neurocomputing: An Introduction - High Dimensional Neurocomputing: Growth, Appraisal and Applications - page 1

Chapter 1
Neurocomputing: An Introduction
Abstract Human brain is the gateway to the deepest mystery of modern computing
science, which confines all the trump cards in the final frontier of technical and sci-
entific inventions. It is an enigmatic and quaint field of investigation having long
past and was always being a remarkable interesting area for researchers. The cru-
cial characteristic of human intelligence is that it is evolving (developing, revealing,
and unfolding) through genetically 'wired' rules and experience. Neurocomputing
is the branch of science and engineering, which is based on human like intelligent
behaviors of machines. It is a vast discipline of research that mainly includes neuro-
science, machine learning, searching and knowledge representation. The traditional
rule-based learning is now appears to be inadequate for various engineering appli-
cations because it is incompetent to serve increasing demand of machine learning
when dealing with large amount of data. This opened up new avenues for the non-
conventional computation models for such applications. Hence, it gives rise to new
area of research, which is named as computational neuroscience. This chapter argues
that we need to understand evolution of information processing in the brain and then
use these principles when building intelligent machines through high dimensional
neurocomputing.
1.1 Evolution of Computational Neuroscience
At first glance, the history of the development of computational neuroscience reveals
the fact that the basic idea of developing the architecture in a way as to ape the neu-
ronal arrangement in the brain came from a study of the neuroscience, brain anatomy,
and analysis of the micro-structure of it. The distinguishing feature that a neuron cell
displayed when impinged with a driving force (voltage) resembled the well-known
sigmoid function that in a study down the line became accepted as the function
of activation of artificial neuron, Fig. 2.1 . The neurocomputing from the incipient
stages grew as more facts about the human brain surfaced with investigation in
Brain Research (Rose and Bynum 1982). In order to process both data and knowl-
edge, neurocomputing evolves its structure (connections and neurons) and function-
ality in a continuous, self organized, online, adaptive and interactive way through