Re: PCA - separation and variance

shay wrote:
I observed (by eye) that a
certain set of variables gives me a good separation using the first
three PCs.

Can you please explain what you mean by separation?

Can I say that the
separation on the higher dimensioned plane is as good as or perhaps
better/worse than the one in the 3D plane?

If it helps, one very general suggestion is that you can reconstruct
correlations between variables based only on the first three PCs and
then compare these to the orginal correlations.

--
John Uebersax PhD

.

Relevant Pages

  • Re: PCA - separation and variance
    ... certain set of variables gives me a good separation using the first ... The more PCs you have, ... That can be found comparing the squared eigenvalues of PCs against the sum of all squared eigenvalues. ... Leave a part of the data out before PCA and try to separate them. ...
    (sci.stat.math)
  • Re: PCA - separation and variance
    ... I observed (by eye) that a ... and that this clustering is detectable by some ... 3D to the actual separation in the higher dimensioned plane? ...
    (sci.stat.edu)
  • Re: PCA - separation and variance
    ... Use discriminant analysis here instead of PCA. ... Then you won't have to determine by eye which variables give you a good separation. ... 3D to the actual separation in the higher dimensioned plane? ...
    (sci.stat.consult)
  • PCA - separation and variance
    ... I have a multi-dimensional data set containing two different groups of ... I observed (by eye) that a ... certain set of variables gives me a good separation using the first ... 3D to the actual separation in the higher dimensioned plane? ...
    (sci.stat.consult)
  • PCA - separation and variance
    ... I have a multi-dimensional data set containing two different groups of ... I observed (by eye) that a ... certain set of variables gives me a good separation using the first ... 3D to the actual separation in the higher dimensioned plane? ...
    (sci.stat.edu)
Loading