Fig. 1
From: Searching for universal model of amyloid signaling motifs using probabilistic context-free grammars
Performance comparison of grammar learning schemes. Mean average precision with regard to mean training duration achieved for the CaMn (a, b) and HET-s (c, d) motifs in the cross-validation experiment. Performance on training and validation sets without smoothing (a, c) and on validation set with smoothing (b, d) is shown. Data for evolutionary learning without (with) contact constraints are plotted in shades of orange (green). Data for statistical learning without (with) contact constraints are plotted in shades of blue (magenta). Note logarithmic scale for duration of training