Is apache hama suitable for implementing adaboost alghoritm?

Question

I'm interested in implementing adaboost algorithm in hadoop environment. I've made research that mapreduce could be slow due to lack of native iterative support. Apache hama is interesting alternative but is there any feature of apache hama which doesn't allow to implement adaboost efficently?

Answer 1

You can implement Adaboost with Hama and BSP. You can also try out Spark, both solutions might work well for Adaboost.