Figure 6
From: Cost function for low-dimensional manifold topology assessment
Cost function applied to PCA projections of experimental Sandia flame data. (a) A 2D projection of flame D using PCA with Max scaling colored by temperature. (b) The overall cost, \(\mathcal {L}\), computed as the L\(_1\)-norm over the individual costs for selected dependent variables: temperature and six important chemical species mass fractions when using PCA for reduction between one and ten dimensions. The figure includes curves for using Max and \(\langle 0, 1 \rangle\) scalings with PCA as well as the cost for the original ten-dimensional manifold.
