Table 1 Results of the ANN models with EGS variables (PTML 3).

From: PTML models of self assembled ligand free nanoparticle catalysts for cross coupling reactions

Model*

R (train)

R (test)

MAE (train)

MAE (test)

RMSE (train)

RMSE (test)

View full size image

0.795

0.822

9.500

9.539

15.353

15.554

View full size image

0.810

0.849

8.385

8.707

14.248

14.560

View full size image

0.778

0.790

15.276

15.347

22.494

22.573

View full size image

0.783

0.859

7.904

8.482

13.802

15.335

View full size image

0.865

0.862

7.383

7.392

12.187

12.252

View full size image

0.917

0.915

5.700

5.864

9.715

9.789

View full size image

0.915

0.915

5.509

5.761

9.817

9.768

View full size image

0.838

0.830

13.261

13.256

19.941

19.894

  1. aModels with the complete dataset. bModels without the double carbonylation subset. cThe codes are BP = Back Propagation, CG = Conjugate Gradient Descent, SS = Subsample, KM = K-Means (Centre Assignment), KN = K-Nearest Neighbour (Deviation Assignment), and PI = Pseudo-Invert (Linear Least Squares Optimization). *Model designations follow the format A:B-C-D-E:F, where A is the number of input features; B, C, and D are the number of neurons in each hidden layer; and E, F are the number of output neurons.
Back to article page