Description

when I use 'mae' criterion for the model extratreesregressor, training for a long time, it's seems lead to an endless training. there have no problem for mse
I find not only me hava this problem. I hava tried two version (0.18 and 0.19.X) , but no used .

https://www.kaggle.com/c/allstate-claims-severity/discussion/24293

Steps/Code to Reproduce

from sklearn.ensemble import ExtraTreesRegressor
rfr = ExtraTreesRegressor(n_estimators=100,
max_features=0.8,  
criterion='mae', 
max_depth=6, 
min_samples_leaf=200,
 n_jobs=-1,
 random_state=17, 
verbose=0)
 mod = rfr.fit(train[distilled_features], train['y'])

Expected Results

can normal training in my model when mae used

Actual Results

aways traing in fit step

Versions

Darwin-16.1.0-x86_64-i386-64bit
Python 3.6.1 |Anaconda 4.4.0 (x86_64)| (default, May 11 2017, 13:04:09)
[GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.57)]
NumPy 1.12.1
SciPy 0.19.0
Scikit-Learn 0.19.X