1. According to the doc of PAMR, the shrinkage
threshold is determined by cross-validation. Does this
mean that user need not tune any parameter?
2. I tried two applications using PAMR, the results
are very disappointing. The attached are the
cross-validation results. You can see that the
classification errors are relatively high (0.2 at the
best), in the case of two categories classification,
the specificity is about 85%, but the sensitivity is
only 55%.
3. From the attched two graphs, we can see that the
shrinkage threshold (varies from 0 to 2) has almost no
effect on the classification error.
Could you suggest how to improve this? or
Does this mean that this method is not suitable for
these two applications?
Many thanks,
luk
---------------------------------
Have fun online with music videos, cool games, IM & more. Check it out!
[[alternative HTML version deleted]]
