Fig. 2.

Comparison between k-spectrum and k-spectrum + shape models on uPBM datasets. (a) Percent of DREAM5 TF datasets that have higher R² values using the k-spectrum + shape model than using the k-spectrum model. (b) Differences of R² values between the two models. (c) The R² performance scores of various k-spectrum models (dashed line) and k-spectrum + shape models (solid line), for $k=1,\dots ,7$