Summary of 3_Default_Xgboost

Extreme Gradient Boosting (Xgboost)

n_jobs: -1
objective: multi:softprob
eval_metric: mlogloss
eta: 0.075
max_depth: 6
min_child_weight: 1
subsample: 1.0
colsample_bytree: 1.0
num_class: 4
explain_level: 0

Validation

validation_type: split
train_ratio: 0.75
shuffle: True
stratify: True

Optimized metric

logloss

Training time

17.3 seconds

Metric details

	1	2	3	4	accuracy	macro avg	weighted avg	logloss
precision	0.738655	0.393701	0.517142	0.664934	0.677055	0.578608	0.665338	0.736726
recall	0.819267	0.0763359	0.458727	0.654	0.677055	0.502083	0.677055	0.736726
f1-score	0.776875	0.127877	0.486186	0.659422	0.677055	0.51259	0.66722	0.736726
support	10391	655	3913	9974	0.677055	24933	24933	0.736726

Confusion matrix

	Predicted as 1	Predicted as 2	Predicted as 3	Predicted as 4
Labeled as 1	8513	4	313	1561
Labeled as 2	86	50	382	137
Labeled as 3	473	56	1795	1589
Labeled as 4	2453	17	981	6523

Learning curves

Learning curves