Table 1 Minimum LOO error on the Leukemia training set (composed of 38 examples), error on the test set (34 examples) and minimum LOO error on the whole Leukemia data set (72 examples).
From: Regularized Least Squares Cancer Classifiers from DNA microarray data
SVM | RLS | |
|---|---|---|
LOO error on training set | 1 | 1 |
Test error | 1 | 1 |
LOO error on the whole data set | 1 | 1 |