A Novel Material Testing Method for Signal Classification

Abstract:

In this paper an unsupervised learning algorithm is proposed. We use Independent Component Analysis algorithm to extract the most essential and highly independent features of video materials. It can assist in testing samples and exacting unknown data to obtain favorable classification goal, which is to obtain a good learning ability for generalization. Experiment results show that the approaches are effective to improve the classification accuracy

• Periodical: Advanced Materials Research (Volume 487)
• Pages: 153-156
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.487.153
• Online since: March 2012
• Authors: Jian Xiang
• Keywords: Face Detection, Machine Learning (ML), Video

References

[1] D Zhao, Z Lin, R Xiao, et al, Linear laplacian discrimination for feature extraction, Proceedings of CVRP, 2007.

[2] J.Proriol, Selection of variables for neural network analysis, Journal of Nuclear Instrument and Methods, 1996, 361:581-585

[3] M.A. Carreira, Z, Lu, Dimensionality Reduction by Unsupervised Regression.Proceedings of CVPR 2008.

[4] Proriol J, Selection of variables for neural network analysis, J. Nucl. Instr. And Meth, 1996, (361):581-585

[5] Ze Lin, Gang Hua, Larry Davis, Multiple Instance Feature for Robust Part-based Object Detection, Proceedings of CVPR 2009: 405-412.(2009)

DOI: 10.1109/cvpr.2009.5206858

[6] T. Kon, K.H. Lee, J. Lee, S. Takahashi, Group Motion Editing, Proceedings of SIGGRAPH, (2008)