I just run the RF examples, non-distributed version: BreimanExample with glass data, 10 iterations with 100 trees, here is the unexpected output:
13/04/25 15:38:40 INFO df.BreimanExample: ******************************************** 13/04/25 15:38:40 INFO df.BreimanExample: Random Input Test Error : 1.0 13/04/25 15:38:40 INFO df.BreimanExample: Single Input Test Error : 1.0 13/04/25 15:38:40 INFO df.BreimanExample: Mean Random Input Time : 0h 0m 1s 79 13/04/25 15:38:40 INFO df.BreimanExample: Mean Single Input Time : 0h 0m 0s 845 13/04/25 15:38:40 INFO df.BreimanExample: Mean Random Input Num Nodes : 6719 13/04/25 15:38:40 INFO df.BreimanExample: Mean Single Input Num Nodes : 11260 is this a mahout bug? or just an unverified implementation? or as the name indicating, 'random forest' just give random predictions? anyone knows any verified random forest java code available online? thanks,