Hi,

I used scikit-learn's LDA for dimensionality reduction and noticed that the 
projected linear discriminants look a little bit strange. For testing, I then 
used the Iris dataset and could reproduce the results that are posted on 
scikit-learn's example documentation here: 
http://scikit-learn.org/stable/auto_examples/decomposition/plot_pca_vs_lda.html

However, when I implemented the LDA in a step by-step approach, the results 
look quite different and more like somehing I would expect. I tried 3 different 
approaches
 
1) on the raw data
2) on z-score standardized data (although all units are centimeter in the Iris 
dataset)
3) on mean-centered data

The results from 2) and 3) are similar to what R's LDA function produces as a 
hint that there might be no error in my step-by-step implementation.

The code that I used for the comparison can be found here: 
http://scikit-learn.org/stable/auto_examples/decomposition/plot_pca_vs_lda.html

It would be great if someone could have a look at this issue. I posted it as 
question to StackOverflow with images to explain the observed differences: 
http://stackoverflow.com/questions/25003427/resulting-projection-from-scikit-learns-lda-is-quite-different-from-lda-in-r-or
------------------------------------------------------------------------------
Infragistics Professional
Build stunning WinForms apps today!
Reboot your WinForms applications with our WinForms controls. 
Build a bridge from your legacy apps to the future.
http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to