Skip to main content

Advertisement

Figure 2 | BMC Microbiology

Figure 2

From: Predicting transcriptional regulatory interactions with artificial neural networks applied to E. coli multidrug resistance efflux pumps

Figure 2

Classification results of the feed-forward (FF) and bi-fan (BF) neural network predictors. Error, precision, recall and f-measure rates were measured for the six different feature vector types, namely Pearson (p), Spearman (s), Kendall (k), partial correlation (pc), Spearman/Kendall/Pearson (skp), and another type containing all previous measures (all). Hybrid models (skp and all) outperformed configurations using only one type of correlation (see analysis in the text). All rates represent the average value over the 100 iterations of the 10 × 10-fold cross-validation procedure.

Back to article page