Fig. 6From: binomialRF: interpretable combinatoric efficiency of random forests to identify biomarker interactionsCalculating RF features’ interactions. a 2-way Interactions. To extend the binomialRF algorithm for 2-way interaction selection, we define the test statistic which reflects the frequency, Fij of the pair Xi ⊗ Xj occurring in the random forest. In particular, the probability of an interaction term occurring by random chance is recalculated and normalized by a factor of a half. b K-way interactions, K = 4. Here, we illustrate the tree traversal process to identify all 4-way interactions, \( \otimes {\boldsymbol{X}}_{i=1}^4 \), with each color denoting a possible interaction path. The legend on the right shows how each interaction path results in a set of 4-way feature interactions. In general, for any user-desired K, the k.binomialRF algorithm traverses the tree via dynamic tree programming to identify all possible paths from the K-terminal nodes to the root, where K-terminal nodes are all nodes K-steps away from the root nodeBack to article page