Figure 11
From: Clustering gene expression data with a penalized graph-based metric
Results on gene expression datasets: Comparison with other state-of-the-art clustering methods (Correlation base metric) Evaluation of the PKNNG metric on eight gene expression datasets using Pearson's correlation as base metric. For each dataset we compare our proposal (the PKNNG metric plus PAM or HC) with other four methods: Model-Based Clustering (MBC), Spectral clustering (Spectral), Evidence accumulation with average linkage (EAC-av) and the Multi-K method.