[MLlib] How to set Loss to Gradient Boosted Tree in Java

[diplomatic Guru]

Hello guys,

I think the default Loss algorithm is Squared Error for regression, but how
do I change that to Absolute Error in Java.

Could you please show me an example?