Skip to main content
Figure 3 | BMC Systems Biology

Figure 3

From: Hierarchical Cluster-based Partial Least Squares Regression (HC-PLSR) is an efficient tool for metamodelling of nonlinear dynamic models

Figure 3

Selection of the optimal number of clusters in HC-PLSR for the gene regulatory network motifs. A) R2 and B) RMSEP values from calibration set predictions of the gene regulatory network time series using from 1 to 10 clusters in HC-PLSR. The calibration set observations were first used for calibration of the HC-PLSR model and subsequently treated as "new observations" and classified prior to the predictions. Each line represents one gene regulatory network motif. The X-axes are truncated since RMSEP increases enormously when the number of clusters is larger than 3. One cluster is equivalent to ordinary PLSR.

Back to article page