Acceso abierto

Artificial Neural Network and Regression Models to Evaluate Rheological Properties of Selected Brazilian Honeys


Cite

Fig. 1

RMSE of training (solid line) and test (dotted line) sets versus number of iterations for optimum MLP ANN: a) model 1; b) model 2; c) model 3; d) model 4.
RMSE of training (solid line) and test (dotted line) sets versus number of iterations for optimum MLP ANN: a) model 1; b) model 2; c) model 3; d) model 4.

Statistical Indexes of Input and Output data in the training and test process of Multilayer Perceptron Feedforward Neural Network

ModelTraining dataTest dataTotal data
MeanSTDMinMaxMeanSTDMinMax
1(*)InputsWC (%)15.781.0214.2318.8615.781.0314.2318.86320
T (°C)30.9015.8810.0060.0032.3117.8210.0060.00
Outputsη (Pa.s)24.9236.910.37225.2727.1738.730.23177.99
2(*)InputsWC (%)15.821.0314.2318.8615.750.9914.2318.86800
T (°C)39.2821.423.6774.5739.2221.833.6174.48
OutputsG′ (Pa)1.273.560.0036.671.182.760.0020.82
G″ (Pa)299.97744.950.926174.37308.74676.121.434571.52
η* (Pa.s)47.74118.560.15982.7049.14107.610.23727.59
3(*)InputsWC (%)15.801.0314.2318.8615.730.9914.2318.86800
T (°C)35.9621.350.5571.4335.9521.710.5771.42
G′ (Pa)28.2970.770.00772.1322.8548.720.00354.15
OutputsG″(Pa)493.781169.190.838455.12530.611282.021.028398.52
η* (Pa.s)78.93186.420.131346.1484.66204.160.161336.82
4(**)WC (%)15.781.0214.2318.8615.791.0214.2318.864160
InputsT (°C)31.0816.2410.0060.0031.7616.6710.0060.00
F (Hz)2.422.950.1010.002.342.890.1010.00
G′ (Pa)2.428.290.00233.522.297.140.00151.11
OutputsG″ (Pa)407.861081.780.1816404.51405.201052.590.239751.40
η* (Pa.s)27.7440.980.28268.3226.9439.960.28266.10

RMSE and correlation coefficient (r) of model 4 variables from the development and test process of a multiple-second order polynomial regression

VariableModel Coefficient1Training dataTest data
β0β1β2β3β12β13β23β123β11β22β33RMSErRMSEr
G′ (Pa)-0.01−0.040.10−0.02−0.17−0.170.26-0.050.060.02950.61890.01700.6932
G″ (Pa)0.03−0.03−0.200.38-−0.40−0.510.560.040.20-0.03620.85030.02930.8539
η* (Pa.s)0.49−0.53−1.29-0.52---0.160.81-0.06800.89640.06400.9018

RMSE and correlation coefficient (r) of models 1, 2 and 3 variables from the development and test process of a nonlinear exponential and of models 1, 2, 3 and 4 from the best ANNs models

ModelEstimated variableEmpirical constants1Exponential Model (Training data)Exponential Model (Test data)ANN (Training)ANN (Test)
ABCRMSErRMSErRMSErRMSEr
1η (Pa.s)0.7992.0366.4150.01010.99810.04330.97000.03590.97600.04300.9681
G′ (Pa)0.8053.30214.6960.03430.93560.02900.92930.03380.93980.02610.9390
2G″ (Pa)0.8812.54910.9530.02820.97250.02720.96880.02960.97040.02520.9731
η* (Pa.s)0.8812.54910.9530.02820.97250.02720.96880.02950.97050.02520.9731
G′ (Pa)0.4892.6288.5420.06590.69480.05210.70550.06750.69690.04860.6629
3G″ (Pa)0.9111.68412.5490.03130.97420.03200.97790.03080.97580.03260.9794
η* (Pa.s)0.9131.68712.5260.03140.97420.03200.97790.03090.97590.03260.9794
G′ (Pa)-------0.02600.72440.01580.7301
4G″ (Pa)-------0.01950.96040.01760.9581
η* (Pa.s)-------0.04200.96360.04060.9647
eISSN:
2299-4831
Idioma:
Inglés
Calendario de la edición:
2 veces al año
Temas de la revista:
Life Sciences, Zoology, other