Skip to main content

Table 2 Classification of lean and obese subjects

From: Gene expression profiling in whole blood identifies distinct biological pathways associated with obesity

  True Positive Rate (Sensitivity) False Positive Rate True Negative Rate (Specificity) False Negative Rate
Classifier Full Filtered Full Filtered Full Filtered Full Filtered
Naïve Bayes 0.96 (0.09) 0.93 (0.08) 0.18 (0.16) 0.003 (0.02) 0.82 (0.17) 0.99 (0.02) 0.03 (0.09) 0.06 (0.08)
Logistic Regression 0.98 (0.04) 0.95 (0.07) 0.11 (0.16) 0.01 (0.03) 0.89 (0.16) 0.99 (0.03) 0.01 (0.04) 0.05 (0.08)
Random Forests 0.95 (0.09) 0.94 (0.11) 0.17 (0.17) 0.10 (0.14) 0.83 (0.17) 0.89 (0.14) 0.05 (0.09) 0.06 (0.11)
ZeroR 0.81 (0.39) 0.81 (0.39) 0.81 (0.39) 0.81 (0.39) 0.19 (0.39) 0.19 (0.39) 0.19 (0.39) 0.19 (0.39)
  1. Comparison of classifier performance for predicting obese and lean phenotype with a full (183 genes) or filtered (11 gene) gene inputs