Skip to main content

Table 1 KL Divergence on Training, Cross Validation and Testing Set

From: Exploiting and integrating rich features for biological literature classification

Unigram Feature Relevant Probability Irrelevant Probability
Training Set Vs Cross Validation Set 0.0216 0.0703
Training Set Vs Testing Set 0.0369 0.9926
  1. (Top 50 features according to Chi-Square statistics)