Abstract
t-Pattern classification is data mining task which maps
data into predefined groups or textit classes. It comes under
supervised learning because the classes are determined before
examining the data. All approaches to performing classification
assume some knowledge of the data. For this a training set is
used to develop the specific parameters required. The problems
of prediction or classification can be solved by using neural
networks (NN). An NN can be said to be a data processing
system, consisting of a large number of simple, highly
interconnected processing elements called as artificial neurons,
in an architecture inspired by the structure of the cerebral
cortex of the brain. The interconnected neural computing
elements have the quality to learn and thereby acquire
knowledge and make it available for use. In present work, three
training algorithms of artificial feed forward neural networks
namely: Back propagation Algorithm, Modified Back
propagation Algorithm and Optical Back propagation
Algorithm are compared on the basis of their error functions.