Neurocomputing: An Introduction - High Dimensional Neurocomputing: Growth, Appraisal and Applications - page 5

won the Nobel Prize for developing Leucotomy to treat psychoses. Walter Rudolf
Hess won Nobel Prize for work on 'Interbrain.' In 1953 A.D., Eugene Aserinski and
Nathaniel Kleitman described the rapid eye movements (REM) during sleep. John
Carew Eccles, Alan Lloyd Hodgkin, and Fielding Huxley shared Nobel Prize for
work on the mechanisms of the neuron cell membrane. Ragnar Arthur Granit, Halden
Keffer Hartline, and George Wald shared Nobel Prize for work on the mechanisms
of vision. Konrad Z. Lorenz, Nicolas Tinbergen, and Karl von Frisch shared Nobel
Prize for work on Ethology. Roger Guillemin and Andrew Victor Schally shared
Nobel Prize for work on peptides in the brain in 1977 A.D. David Hunter Hubel
and Torsten N. Wiesel shared Nobel Prize for explaining the visual system in 1981
A.D. In the same year, Roger Walcott Sperry won a Nobel Prize for explaining the
functions of the cerebral hemispheres. In 1982 A.D., Bengt Ingemar Bergstrom,
John Robert Vane, and Sune K. Bergstrom won Nobel Prize for the discovery of the
Prostaglandins. In 1982 A.D., Stanley Cohen and Rita Levi-Montalcini won Nobel
Prize for their work on the control of nerve cell growth. In 1991 A.D., Erwin Neher
and Bert Sakmann shared Nobel Prize for their work on the function of single ion
channels. In 1994 A.D., Alfred G. Gilman andMartin Rodnell shared Nobel Prize for
their discovery of G-Protein coupled receptors and their role in signal transduction.
For their discoveries concerning signal transduction in the nervous system Arvid
Carlsson, Paul Greengard and Eric Kandel shared Nobel Prize in 2000 A.D.

1.1.2 Neurocomputing Research Chronology

Neurocomputing inspired from biological neurons endeavors to achieve human-like
ability of intelligence for learning and generalization. Neurocomputing-based tech-
niques have better ability to deal with nonideal scenarios, therefore, they can provide
robust and efficient solutions to such situations. Computing derived from the typical
convergence of intelligent techniques focuses natural way of problem solving with
primarily inclusion of neural network. The mathematical formulation about the con-
vergence of composition of functions was first time demonstrated by Kolmogorov
(1957) who stated that any continuous real valued functions in n—variables defined
on [0 1] n , ( n > 2 ) can be represented in the form of composition of function as

shown

2 n + 1 n

ψ( x 1 , x 2 ,... x n ) = f j g ij ( x i ) (1.1)

j = 1 i = 1

When the function was proposed, it was not apparent how and where it could
be applied which left the investigators of the time baffled. Many results followed
the discovery of the ANN that stated how and why the neural network algorithms
would converge. For instance, there were results on how to select activation func-
tions to ensure the neural network scheme converged (Hassoun 1995). Some efforts