Receiver Operating Characteristic (ROC) curves and the Precision Recall Curve both demonstrate the performance of the SPACE algorithm on the 231-gene network with 20 samples and a noise value of 0.25 when performing a single iteration (<i>i.e.</i>, “non-bootstrapped”) or bootstrapping the dataset using the Ensemble Network Aggregation approach.

<p>In this case, the Area Under the ROC Curve (AUC) of the non-bootstrapped SPACE method was 0.748, while that of the bootstrapped SPACE method was 0.816. The Area Under the Precision-Recall (AUPR) curve also improves from 0.249 (SPACE) to 0.273 (bootstrapping).