Bayes PCA Revisited

Jon SporringResearch 2008 · OpenAlex

Abstract

<p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">Principle Component Analysis is a simple tool to obtain linear models for <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">stochastic data and is used both for a data reduction or equivalently noise elim- <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">ination and for data analysis. Principle Component Analysis ts a multivariate <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">Gaussian distribution to the data, and the typical method is by using the log- <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">likelihood estimator. However for small sets of high dimensional data, the log- <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">likelihood estimator is often far from convergence, and therefore reliable models <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">must be obtained by use of prior information. In this paper, we will examine <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">an earlier work on reconstructing missing data using statistical knowledge and <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">regularization, we will show the circumstances for which this is equivalent to <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">a Bayes estimation, we will give an expository presentation of Bayes Principle <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">Component Analysis for a range of exponential type priors, and we will develop <p style="margin: 0px; font-family: Helvetica; font-style: normal; font-variant: normal; font-weight: normal; font-size: 9px; line-height: normal; font-size-adjust: none; font-stretch: normal">algorithms for their estimate.
BibTeX
@inproceedings{sporring2008bayes,
  author = {Jon Sporring},
  title = {{Bayes PCA Revisited}},
  booktitle = {Research},
  year = {2008},
}

Loading related content...

Figure