Figure 5
From: Learning smoothing models of copy number profiles using breakpoint annotations
Training error, false positive (FP) and false negative (FN) were calculated for all 17 algorithms applied to the original and the detailed annotation databases. For each profile and algorithm, the smoothing parameter with minimal breakpoint annotation error was selected, and we report the mean training error across all profiles. Squares show the same colors as in the figures, and are absent for default models that have no smoothness parameters.