Table 1 Comparison statistics of multilayer perceptron (MLP) and radial basis function (RBF) for various morphological traits of pomegranate.

From: Artificial neural network-based model to predict the effect of γ-aminobutyric acid on salinity and drought responsive morphological traits in pomegranate

Model

Subset

Criterion

CD

PH

LLI

LWI

LAI

MLP

Training

R2

0.88

0.95

0.76

0.79

0.97

RMSE

0.46

2.1

0.30

0.11

0.47

MBE

− 2.77

− 0.0001

− 6.33

0.0001

− 5.57

Testing

R2

0.76

0.89

0.83

0.8

0.96

RMSE

0.66

3.04

0.35

0.13

0.5

MBE

− 0.02

− 0.34

− 0.04

0.0003

0.006

RBF

Training

R2

0.71

0.83

0.84

0.83

0.87

RMSE

0.7

3.82

0.31

0.11

0.98

MBE

− 8.99

0.0002

− 1.9

8.83

0.0001

Testing

R2

0.74

0.86

0.74

0.84

0.85

RMSE

0.8

4.16

0.38

0.13

1.08

MBE

0.22

0.3

0.0001

0.01

0.07

  1. PH plant height, CD crown diameter, LLI leaf length index, LWI leaf width index, LAI leaf area index, R2 coefficient of determination, RMSE root mean square error, MBE mean bias error.
Back to article page