Artificial neural network modeling of a multiphase photodegradation system

Predictive Abilities of Bayesian Regularization and Levenberg–Marquardt Algorithms in Artificial Neural Networks: A Comparative Empirical Study on Social Data

“…where Ij is the relative importance of the input factors j for the output, n is the number of input factors, h is the number of hidden neurons, W is the synaptic weight matrix between the input and the hidden layer, and WO is the synaptic weight matrix between the hidden and output layers [28,30]. Table 5 and Figure 3 show the relative importance of 24 input neurons upon the target for BR models.…”

Comparison of Response Surface Methodology and Artificial Neural Network in Predicting the Microwave-Assisted Extraction Procedure to Determine Zinc in Fish Muscles

“…A biological neuron receives inputs from other sources, combines them, generally performs a non-linear operation on the result, and then outputs the final result [6]. The ability of the ANNs, to recognize and reproduce the cause-effect relationships through training for the multiple input-output systems makes them efficient to represent even the most complex systems [7]. The main advantages of ANN as compared to RSM include: 1) ANN does not require any prior specification of suitable fitting function, and 2) ANN also has a universal approximation capability to approximate almost all kinds of non-linear functions including quadratic functions, whereas RSM is useful only for quadratic approximations [8].…”

Applications of Neural Networks in Advanced Oxidative Process

“…In recent years, neural networks have been applied in various areas in the Chemical Engineering and, concerning the Advanced Oxidation Process it can be quoted the work of Pareek et al (2002) in which it was studied the photodegrading of Spent Bayer liquor, with the use of a feedforward-type neural network. Pearson correlation coefficients above 0.99 were obtained in this work.…”