PCA stability and choice of dimensionality.

*(English)*Zbl 0743.62046Summary: A criterion of stability for PCA scatterplots is defined based on a classical distance between projectors. It is constructed as a risk function and can be estimated by bootstrap or jackknife methods. Furthermore, perturbation theory is used to write down a Taylor expansion of the jackknife estimate for reasons of computational cost and in order to obtain an analytic expression for the approximation. The comparative study of these three estimates on real data shows that the last one is easy to compute, sufficiently accurate and helpful in choosing dimensionality in PCA.

##### MSC:

62H25 | Factor analysis and principal components; correspondence analysis |

62-09 | Graphical methods in statistics (MSC2010) |

##### Keywords:

principal components analysis; PCA scatterplots; classical distance between projectors; risk function; bootstrap; perturbation theory; Taylor expansion; jackknife estimate; dimensionality##### Software:

R
PDF
BibTeX
XML
Cite

\textit{P. Besse}, Stat. Probab. Lett. 13, No. 5, 405--410 (1992; Zbl 0743.62046)

Full Text:
DOI

##### References:

[1] | Becker, R.A.; Chambers, J.M.; Wilks, A.R., The new S language, a programming environment for data analysis and graphics, (1988), Wadsworth & Brooks/Cole Pacific Grove · Zbl 0642.68003 |

[2] | Beran, R.; Srivastava, M.S., Bootstrap tests and confidence regions for functions of a covariance matrix, Ann. statist., 13, 95-115, (1985) · Zbl 0607.62048 |

[3] | Chatelin, F., Valeurs propres de matrices, (1988), Masson Paris · Zbl 0691.65018 |

[4] | Critchley, F., Influence in principal components analysis, Biometrika, 72, 627-636, (1985) · Zbl 0608.62068 |

[5] | Daudin, J.J.; Duby, C.; Trécourt, P., Stability of principal components analysis studied by the bootstrap method, Statistics, 19, 241-258, (1988) · Zbl 0643.62043 |

[6] | Daudin, J.J.; Duby, C.; Trécourt, P., P.C.A. stability studied by the bootstrap and the infinitesimal jackknife method, Statistics, 20, 255-270, (1989) · Zbl 0671.62038 |

[7] | Dauxois, J.; Pousse, A.; Romain, Y., Asymptotic theory for the principal component analysis of a vector random function: some applications to statistical inference, J. multivariate anal., 12, 136-154, (1982) · Zbl 0539.62064 |

[8] | Efron, B., The jackknife, the bootstrap and other resampling plans, (1982), SIAM Philadelphia, PA · Zbl 0496.62036 |

[9] | Gauss, The Gauss system version 2.0, (1988), Aptech Systems Kent |

[10] | Jolliffe, I.T., Principal component analysis, (1986), Springer New York · Zbl 1011.62064 |

[11] | Kato, T., Perturbation theory for linear operators, (1966), Springer New York · Zbl 0148.12601 |

[12] | McDonald, G.C.; Ayers, J.A., Some applications of the “chernoff faces’: A technique for graphically representing multivariate data, () |

[13] | Winsberg, S., Two techniques: monotone spline transformations for dimension reduction in PCA and easy to generate metrics for PCA of sampled functions, () |

This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.