Issues in the Use of Neural Networks in Information Retrieval

Abstract

The series "Studies in Computational Intelligence" (SCI) publishes new developments and advances in the various areas of computational intelligence-quickly and with a high quality. The intent is to cover the theory, applications, and design methods of computational intelligence, as embedded in the fields of engineering, computer science, physics and life sciences, as well as the methodologies behind them. The series contains monographs, lecture notes and edited volumes in computational intelligence spanning the areas of neural networks, connectionist systems, genetic algorithms, evolutionary computation, artificial intelligence, cellular automata, self-organizing systems, soft computing, fuzzy systems, and hybrid intelligent systems. Of particular value to both the contributors and the readership are the short publication timeframe and the worldwide distribution, which enable both wide and rapid dissemination of research output.

Dealing the Nonlinearity Associated With the Data Using Artificial Neural Networks

“…Example 1. We are interested [6] in exploring for a sample of 32 vehicles the relationship between: the number of gears of a vehicle, the overall length (in inches) and its fuel efficiency (measured in miles per gallon).  the number of the classes (the number of the neurons belonging to the last layer);  the equations of the fourth layer, but the structure diagram is similar.…”

“…The performance of FGNN over MP is based on [6] the fuzzy properties of FGNN, while the MP is a crisp neural network. The comparison [6] of FGNN and respectively MP versus MLRM marks both the competition nonlinear over linear and of neural over statistical, too.…”

“…Each neuron of this level performs a Gaussian membership function [2], [6], [9]. The FGNN parameters have a physical significance, in the meaning that:…”

“…The connections between the membership neurons of the second layer and the rule neurons that characterize the third layer of the FGNN indicate the premise of the fuzzy rules. This layer computes the antecedent matching by the product operation [2], [6], [9].…”

“…The FGNN parameters one initialize according to the on-line initialization algorithm [2], [6], [9] and they will be refined during the training algorithm [2], [6][7][8][9].…”