Table 1

The average Kullback Leibler divergence between the distributions of different datasets

Compared distributions

Term feature

String feature


Pr(x|c+)

Pr(x|c-)

Pr(x|c+)

Pr(x|c-)


Dist on the remaining training dataset versus Dist on the leave-out dataset

0.0216

0.0703

0.0029

0.0163

Dist on the remaining training dataset versus Dist on the official test dataset

0.0369

0.9926

0.0357

0.1887


The table shows the average Kullback Leibler divergence of three distributions estimated on the leave-out dataset, remaining training dataset, and the official test data. The Average Kullback Leibler divergence between distributions on different datasets. Dist, distribution.

Huang et al. Genome Biology 2008 9(Suppl 2):S12   doi:10.1186/gb-2008-9-s2-s12

Open Data