Table 7 Optimized parameter values for E-BPC.

From: Predicting workability and mechanical properties of bentonite plastic concrete using hybrid ensemble learning

Parameters

Models

ADB-FBIO

GBRT- FBIO

XGB- FBIO

RF- FBIO

Number of estimators

7

44

16

5

Min_samples_split

0.02

1.00e-08

-

2

Min_samples _leaf

0.002

0.003

-

1

Max_depth

227

298

254

499

Max_features

1

2

-

7

Max_ leaf_nodes

309

9

271

168

Ccp_alpha

0.00019

0

-

1.42

Min_weight_fraction_leaf

0.0018

0.02

-

0.0007

Learning rate

0.42

0.59

1.36

-

Alpha

-

0.05

-

-

Subsample

-

0.98

0.85

-

Max_samples

-

-

-

0.97

Gamma

-

-

0.0003

-

Min_child_weight

-

-

0.81

-

Reg_lambda

-

-

1.007

-

Colsample_bytree

-

-

0.88

-

Back to article page