Training ensembles of randomized decision trees

摘要: A method training a randomized decision tree through multiple iterations, each is based on: a) Receiving data samples that include subsets, subset corresponds to an attribute. b) Distributing the subsets slave processing units after sorting in consecutive ascending order by updating first index identifies trajectories of nodes previous level. c) Simultaneously identify split functions for node with respect and second current d) Collecting from constructing level selecting preferred function