The tree learners: optimal (optimal: DL8.5 and depth-2 solvers) and greedy with lookahead (greedy: LGDT).
optimal
greedy