Journal of Statistics: Advances in Theory and ApplicationsAuthor's: Xiangyu Wang
Pages: [21] - [32]
Received Date: January 18, 2017
Submitted by:
DOI: http://dx.doi.org/10.18642/jsata_7100121760
In this paper, a new missing data filling method, SG-PCA filling algorithm, is provided. This algorithm is based on principal component analysis and Savitzky-Golay denoising method. For given incomplete data, we apply the PCA filling algorithm and the Savitzky-Golay denoising method alternatively to approximate the missing values. As an example, a filling experiment is performed by using the Breast Cancer data set in the University of California Irvine (UCI). The results show that the SG-PCA filling algorithm is more effective in filling accuracy.
Savitzky-Golay denoising (SG), principal component analysis (PCA), missing data filling.
