On various multi-layer perceptron and radial basis function based artificial neural networks in the process of a hot flow curve description

In recent years, the study of hot deformation behavior various materials is significantly marked by an increasing utilization artificial neural networks, which are frequently employed for a flow curve description. This specific kind description commonly solved via Feed-Forward Multi-Layer Perceptron architecture and rarely Radial Basis architecture. Both network architectures compared to assess their suitability in process under wide range thermomechanical conditions. The performed survey also aimed on eventual corresponding modifications both studied namely Cascade-Forward Generalized Regression network. main results have shown that represents good choice if very high accuracy crucial goal. However, case this architecture, finding proper parameters can be time-consuming hardware burdensome. On contrary, almost unused offers easy training procedure shorter computing time acceptable accuracy. submitted research should then serve as background selection following application suitable solving future tasks.