Table 8 The outcome of training algorithms for Backpropagation in Multilayer Perceptron.

From: Optimization and modeling of sulfur removal from liquid fuel using carbon-based adsorbents through synergistic application of RSM and machine learning

Backpropagation algorithm

Function

Number of neurons

R2

MSE train

MSE test

Epoch

Levenberg–Marquardt

Trainlm

46

0.980

0.0028

0.0336

32

Bayesian regularization

Trainbr

23

0.975

0.0034

0.0415

48

Scaled conjugate gradient

Trainscg

46

0.910

0.0152

0.0983

23

Back to article page