FACE CLASSIFICATION USING WIDROW-HOFF LEARNING PARALLEL LINEAR COLLABORATIVE DISCRIMINANT REGRESSION (WH-PLCDRC)

Learning based face classification technique is proposed in this paper to improve Xiaochao Qu s linear collaborative discriminant regression classification (LCDRC). The LCDRC helps to find out a discriminant subspace by increasing collaborative between-class reconstruction error and decreasing the w...

Full description

Saved in:
Bibliographic Details
Published in:Journal of Theoretical and Applied Information Technology 2016-07, Vol.89 (2), p.362-362
Main Authors: Vishwanath, P, Viswanatha, V M
Format: Article
Language:English
Subjects:
Classification
Errors
Learning
Methodology
Optimization
Projection
Reconstruction
Regression
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Learning based face classification technique is proposed in this paper to improve Xiaochao Qu s linear collaborative discriminant regression classification (LCDRC). The LCDRC helps to find out a discriminant subspace by increasing collaborative between-class reconstruction error and decreasing the within-class reconstruction error simultaneously. With the aim of proposing the further improvement of the accuracy of probe image classification by minimizing the value of WCRC with the aid of our proposed methodology. Proposed methodology used parallel process to find an optimal projection matrix with the aid of the LCDRC method (arm A) & implement Widrow-Hoff learning method into the LCDRC (arm B). Both arms A & arm B are varying with the calculation of WCRC & arm A returns the current value of WCRC & arm B derives the future value of WCRC. Then, the optimal projection matrix is derived by the proposed methodology after the selection of suitable value of WCRC derived from the arm A & arm B method.
ISSN:1817-3195