Weighted principal component analysis

Authors:
Zizhu Fan;Ergen Liu;Baogen Xu
Affiliations:
School of Basic Science, East China Jiaotong University, Nanchang, Jiangxi, China;School of Basic Science, East China Jiaotong University, Nanchang, Jiangxi, China;School of Basic Science, East China Jiaotong University, Nanchang, Jiangxi, China
Venue:
AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part III
Year:
2011

Citing 10
Cited 1

Principal component neural networks: theory and applications

Principal component neural networks: theory and applications
Linear Discriminant Analysis of MPF for Face Recognition

ICPR '98 Proceedings of the 14th International Conference on Pattern Recognition-Volume 1 - Volume 1
Two-Dimensional PCA: A New Approach to Appearance-Based Face Representation and Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
An improved face recognition technique based on modular PCA approach

Pattern Recognition Letters
Eigenfaces for recognition

Journal of Cognitive Neuroscience
An approach for directly extracting features from matrix data and its application in face recognition

Neurocomputing
A feature extraction method for use with bimodal biometrics

Pattern Recognition
Face Recognition Based on Two Dimension Double PCA and Affinity Propagation

ICNC '09 Proceedings of the 2009 Fifth International Conference on Natural Computation - Volume 05
Manifold learning for video-to-video face recognition

BioID_MultiComm'09 Proceedings of the 2009 joint COST 2101 and 2102 international conference on Biometric ID management and multimodal communication
LPP solution schemes for use with face recognition

Pattern Recognition

Pointwise probability reinforcements for robust statistical inference

Neural Networks

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this paper, we proposed a weighted PCA (WPCA) method. This method first uses the distances between the test sample and each training sample to calculate the 'weighted' covariance matrix. It then exploits the obtained covariance matrix to perform feature extraction. The experimental results show that the proposed method can obtain a high accuracy than conventional PCA. WPCA has the underlying theoretical foundation: through the 'weighted' covariance matrix, WPCA takes emphasis on the training samples that are very close to the test sample and reduce the influence of the other training samples. As a result, it is likely that the test sample is easier to be classified into the same class as the training samples that are very close to it. The experimental results show the feasibility and effectiveness of WPCA.