Numerical study on the feasibility of dynamic evolving neural-fuzzy inference system for approximation of compressive strength of dry-cast concrete

- For the first time, dynamic evolving neural-fuzzy system (DENFIS) was used to predict the compressive strength of dry-cast concretes.
- For comparison purposes, 6 nonlinear regression, 6 neural network, 5 ANFIS, 3 online first-order TSK DENFIS, 3 offline first-order TSK DENFIS, and 3 offline high-order TSK DENFIS models were developed.
- DENFIS model with high-order TSK inference system was found to be more robust than first-order TSK online and offline models.
- High-order DENFIS model could be trained to produce more reliable prediction results in comparison with neural network, ANFIS and nonlinear regression models.