Start with some initial target.
Build a depth one classification tree over left-hand side variables.
Relabel the entities in the leaves of the tree.
Use these new labels as the target.
Build a depth one classification tree over right-hand side variables.
Relabel the entities in the leaves of the tree.
Use these new labels as the target.
Split the target according to the tree under construction for the LHS.
Build depth one classification trees over left-hand side variables separately for either split.
Relabel the entities in the leaves of the tree.
Use these new labels as the target.
Split the target according to the tree under construction for the RHS.
Build depth one classification trees over right-hand side variables separately for either split.
Relabel the entities in the leaves of the tree.
Use these new labels as the target.
Split the target according to the tree under construction for the LHS.
Build depth one classification trees over left-hand side variables separately for each split.
Relabel the entities in the leaves of the tree.
Use these new labels as the target.
Split the target according to the tree under construction for the RHS.
Build depth one classification trees over right-hand side variables separately for each split.
Relabel the entities in the leaves of the tree.
Take the last tree from each side,
select the branches leading to positive leaves,
and compute the supports
and compute the supports across the two sides.