Omatu, Sigeru (Osaka Prefecture Univ.), Fujinaka, Toru (Osaka Prefecture Univ.)

Bank Note Classifier Using Neural Networks

Scheduled for presentation during the Regular Session "Neural Networks II" (ThA12), Thursday, July 27, 2006, 10:25−10:50, Room 104b

17th International Symposium on Mathematical Theory of Networks and Systems, July 24-28, 2006, Kyoto, Japan

This information is tentative and subject to change. Compiled on May 19, 2024

Abstract

This paper addresses the reliability of neuro-classifiers for bank note recognition. A local principal component analysis (PCA) method is applied to remove non-linear dependencies among variables and extract the main principal features of data. At first the data space is partitioned into regions by using a self-organizing map (SOM) model and then the PCA is performed in each region. A learning vector quantization (LVQ) network is employed as the main classifier of the system. By defining a new algorithm for rating the reliability and using a set of test data, we estimate the reliability of the system. The experimental results taken from 1,200 samples of US dollar bills show that the reliability is increased up to 100% when the number of regions as well as the number of codebook vectors in the LVQ classifier are taken properly.