Application of Sequential Learning Neural Networks to Civil Engineering Modeling Problems

Thermodynamically Consistent Vapor-Liquid Equilibrium Modelling with Artificial Neural Networks

“…In simple terms, serial hybrid models generally substitute low-dimensional parametrizations of semiempirical models with ANNs (e.g., an ANN can be used instead of Antoine's equation for computing the pure component saturation pressure). Although not labeled as hybrid models, this approach has been utilized for VLE modelling in the literature by reparametrizing already established semi-empirical models such as NRTL [46], PR EoS [47] or PRSV EoS [48].…”

Accurate identification of geological samples using artificial neural network processing of laser-induced breakdown spectroscopy data

“…3. [23][24][25] Five spectra are obtained by averaging the spectra from each sampling spot of a tablet. These average spectra from all samples are divided into five subsets which contain one spectrum per sample.…”

MicroARTMAP for pattern recognition problems

“…In order to compare the efficiency of MicroARTMAP, the sequential learning neural networks (SLNN) [15] was used with the architecture (7 inputs À 6 input neurons + 1 bias), 1 hidden neuron and 1 output neuron with learning rate of 0.5 and a momentum factor of 0.0000001. It required 10,000 iterations for training.…”