Assume some matrix M with m rows and n columns. Let C(M) denote the covariance matrix computed from M. With principal component analysis, we compute the eigenvectors of C(M) and choose the first k, for example, based on their ordering according to their corresponding eigenvalues. My question is, if e1,e2,...,ek are the eigenvectors that form the principal components, then how can map these back to the original dimensions of M so that I can say that the original dimensions x1,x2,... of M are the ones corresponding to e1,e2,... Is it possible to do this?(adsbygoogle = window.adsbygoogle || []).push({});

**Physics Forums | Science Articles, Homework Help, Discussion**

Join Physics Forums Today!

The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

# PCA mapping back to original dimensions

Can you offer guidance or do you also need help?

Draft saved
Draft deleted

Loading...

Similar Threads for mapping original dimensions |
---|

I External Direct Sums and the Sum of a Family of Mappings ... |

I Linear mapping of a binary vector based on its decimal value |

I Finding the Kernel of a Matrix Map |

**Physics Forums | Science Articles, Homework Help, Discussion**