Table 1 Final R2 scores of the models.

From: Machine learning analysis of rivaroxaban solubility in mixed solvents for application in pharmaceutical crystallization

Model

Number of trees

criterion

Loss function

Learning rate

Max tree depth

Min samples split

GB

180

Friedman MSE

Squared error

0.737607

10

3

LGB

144

Friedman MSE

Squared log error

1.40681454

9

2

ET

153

Absolute error

–

–

9

3

RF

17

Poisson

–

–

13

4

Back to article page