Figure 1
From: LS-NMF: A modified non-negative matrix factorization algorithm utilizing uncertainty estimates
Fits to the Data. The comparison of LS-NMF and NMF for fitting data is shown for the Rosetta Compendium. NMF is based on root mean squared deviation (RMSD) fitting, and as the number of dimensions increases, the fit improves with potential overfitting possible. The χ2 measurement provides the standard measure of fit to the data and shows how the LS-NMF method fits data much better once individual uncertainty measurements are considered.