How do you weight example sets in rapid miner?


How do i weight more recent examples more and historical examples less with rapid miner? For input into a SVM model?
Tagged:
0
0 Comments | 0 Discussions | 0 Members | 0 Online |