Skip to main content

Table 4 8 groups of XGBoost model parameter results

From: Prediction of postoperative infectious complications in elderly patients with colorectal cancer: a study based on improved machine learning

max_depth

min_child_weight

gamma

subsample

colsample_bytree

reg_alpha

reg_lambda

learning_rate

num_boost_round

3

3

0.2

0.9

0.8

1.00E-05

1

0.01

5000

3

3

0.2

0.8

0.8

1.00E-05

1

0.01

5000

3

3

0.4

0.9

0.5

1.00E-05

1

0.01

5000

3

3

0.4

0.95

0.5

1.00E-05

1

0.01

5000

7

4

0

0.9

0.95

1.00E-05

1

0.01

1000

7

4

0

0.95

0.95

1.00E-05

1

0.01

1000

7

4

0.1

0.9

0.3

1.00E-05

1

0.01

1000

7

4

0.1

0.95

0.4

1.00E-05

1

0.01

2000