Fitting algorithms for GRNNS in clustering applications

In this paper a classifier structure applying Generalized Regression Neural Networks to detect microcalifications (μCs) is proposed. It is part of a Computer Assisted Diagnosis (CAD) system designed to detect μCs in digitalized mammographics. Suspicious area of each mammography is selected and store...

Full description

Saved in:
Bibliographic Details
Main Authors: Buendia-Buendia, F.S., Alfaro Rodriguez, J.J., Vega-Corona, A.
Format: Conference Proceeding
Language:English
Subjects:
Clustering algorithms
Intelligent networks
Neural networks
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper a classifier structure applying Generalized Regression Neural Networks to detect microcalifications (μCs) is proposed. It is part of a Computer Assisted Diagnosis (CAD) system designed to detect μCs in digitalized mammographics. Suspicious area of each mammography is selected and stored. A GRNN network classisfies the pixels minimimizing the mean square error (MSE). This structure was selected for its advantegeous features like highly localized pattern nodes and instanteneous learning. Three algorithims to fit the networkd parameters, given a training data set, are proposed, Some guidelines to band up the training data set have been proposed.