Figure 2

Significance of the difference in objective scores. Quadrant plot showing the significance of the difference in objective scores between each normalisation method and unnormalised data. Each normalisation method is represented by a different symbol, global variants being annotated with an additional circle. The axes represent p-values of the statistical analysis described earlier, for classification performance and variance reduction on the x-axis and y-axis, respectively. The point (1, 1) indicates the position of unnormalised data. The region in which a method lies represents its performance for each objective with respect to using no normalisation.