Number Of Codebook Vectors Research Articles

An Enhanced Approach for Classification in Web Usage Mining using Neural Network Learning Algorithms for Supervised Learning

ABSTRACT Now a day data on the web is growing by a rapid speed, large volume of data is available on web. So, extract useful knowledge from large web data, efficient web mining methods are required to handle those data and achieve various functionalities such a as user trend analysis, web profile analysis, Web AD market change analysis, etc. The concept of neural network helps to handle large volume of data by its characteristics [1]. Several neural network learning algorithms provides better supervised learning. They are capable to handle huge dynamic data. Specially, LVQ (Learning Vector Quantization) algorithms are useful for supervised, dynamic labeling or post-training map labeling and supervised version of SOM through that the approximation of distribution of class with less number of codebook vectors and able to minimize classification errors respectively [9]. MLVQ and HLVQ are both the techniques are following a concept of multi pass in which more than one pass can be performed on the same model and is very useful for gaining best desired results [11]. Here in this paper, we are going to discuss a new technique that will work on hierarchical as well as multi pass approach that is having the advantages of both multi pass and hierarchical approach by combining the benefits of both and designed a new algorithm. That new technique will more accurate, less time consuming and able to decrease learning rate for neural network. The basic HLVQ approach will follow the same algorithm for generation of all phases. In addition to that HMLVQ provides better efficiency through enhancing advantages the various approaches for the same classification process.

IMPROVMENT OF BANK NOTE CLASIFICATION RERIBILITY

Abstract 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.

Reliability improvement of bank note classification

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.