search

kzwkt / wnd-charm

Automatically exported from code.google.com/p/wnd-charm
0 stars 0 forks source link

NON-Normalized Similarity Table #16

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1.  Run the current configuration.
2.
3.

What is the expected output? What do you see instead?
Probability table from marginal probability space is normalized.  This works 
well for the current method of calculating distances, however may not be the 
best option for vector subtraction distance calculation.

What version of the product are you using? On what operating system?
WndChrm 1.29

Please provide any additional information below.
Would like the output to contain an additional table with the unmodified 
marginal probability averages / centroids.  

Possibly a switch in the command line that would allow you to use the 
unmodified marginal probability averages to calculate distances based on vector 
subtraction.

Original issue reported on code.google.com by DuncanDe...@gmail.com on 2 Mar 2011 at 6:10

GoogleCodeExporter commented 9 years ago 
Code now in repository.
The default is to compute distances based on class centroids in MP space (-P5). 
 The old way is available via -P2.
Class Probability Matrixes are put in the report per-split as well as in the 
summary.

Original comment by i...@cathilya.org on 3 Mar 2011 at 10:08