Trained hyper-parameters

params = {
   'base_score': 0.5,
   'booster': 'gbtree',
   'colsample_bytree': 0.5,
   'eta': 0.4,
   'eval_metric':  ['error', 'mae'],
   'max_depth': 30,
   'nthread': 4,
   'objective': 'binary:logistic',
   'reg_alpha': 0.001,
   'reg_lambda': 0.4,                                                    
   'scale_pos_weight':8.85,
   'max_delta_step':1,
   'seed': 12345,
   'silent': 0,
   'subsample': 0.5
}

Number of trees: 110

Results evaluation:

best: [109] tr-error:0.02285 tr-mae:0.039291	val-error:0.307175 val-mae:0.330359

trained ../mat_training.train mae: 0.0393 rmse: 0.1355
avg 0.102133 mae: 0.1834 rmse: 0.3028
Predicted correctly: 3921416/4013117

Stats data set:

Training:

ONLY data set generated from veriT(M)

nb_negatives: 1716339 is:88.71055524630148498800 nb_positives: 218424 is: 11.28944475369851501100 nb_examples 1934763
ratio:7,857831557

ALL union data set generated from veriT + veriT(M)

nb_negatives: 3603247 is:89.78674182686425539000 nb_positives: 409870 is: 10.21325817313574460900 nb_examples 4013117
ratio:8,79128

Test:

ONLY data set generated from veriT(M)

nb_negatives: 751098 is:87.96047330847495737200 nb_positives: 102806 is: 12.03952669152504262700 nb_examples 853904

ALL union data set generated from veriT + veriT(M)

nb_negatives: 1718075 is:91.41877014663611703200 nb_positives: 161271 is: 8.58122985336388296700 nb_example 1879346