Skip to main content

Table 5 The relative improvement of MetaRF compared to the baseline method(in Buchwald HTE dataset)

From: MetaRF: attention-based random forest for reaction yield prediction with a few trails

 

RMSE \(\downarrow\)

\(R^{2}\) \(\uparrow\)

Sample\(^a\)

MetaRF

Baseline

Margin\(^b\)

MetaRF

Baseline

Margin\(^b\) (%)

2

14.5912

16.7692

12.99%

0.6986

0.6026

15.94

4

12.7039

16.2212

21.68%

0.7717

0.6277

22.93

6

12.6674

15.3222

17.33%

0.7722

0.6681

15.58

8

12.0527

15.0728

20.04%

0.7929

0.6790

16.77

10

11.8515

14.7419

19.61%

0.7996

0.6922

15.52

  1. aThe number of fine-tune samples
  2. bRelative improvement compared to the baseline method, random forest
Back to article page

Journal of Cheminformatics

ISSN: 1758-2946

Contact us