Figure 1
From: Ciruvis: a web-based tool for rule networks and interaction detection using rule-based classifiers
Rule networks for simulated data. Rule networks for twelve different pairs of maximum correlation X and interaction Y for the “0” outcome. The parameter choices (A-L) correspond to points in Additional file 4: Figure S2. The correlated features are named C0 to C4 (lowest to highest correlation), and the pairs R0, S0 to R4, S4 (lowest to highest correlation). The colors were specified so that the interacting pairs have the same color. Each feature occurs twice in the figure; the first time with the value 0 and the second with 1.