It is worth mentioning that prediction of the concrete compressive strength is an important fact in the quality assurance of the produced concrete. Although there are numerous methods of predicting the mechanical properties of concrete, not all of them are valuable since they are in accordance with many trial and errors. As mentioned earlier, three different models of multiple linear regressing (MLR), artificial neural network (ANN), and ANFIS are used to reach the goal in this study. These models are defined comprehensively in the following sections.
3.1. Multiple Linear Regression Model (MLR) Generally, regression models can be considered as the process of fitting models to data. In order to handle the relationships between the variables, …show more content…
1989, Khademi et al. 2015). Hence, it is highly recommended to use a single hidden layer, and therefore, this recommendation has been used in this paper. Each particular node is a layer is connected to many other nodes using weighted connections. The weights estimate the strength of connections between interconnected nodes. The weights are trained in order to get the objective that the ANN`s output becomes similar to measured output on a recognized data …show more content…
The early stopping method has been known to be the most efficient method for optimizing the generalization performance of ANNs in practice. Therefore, early stopping method can be selected for both increasing the generalization capability of the model and overcoming the over-fitting. Normally, the data in ANN are divided into two sets of training and testing. Using the test subset, the performance of the training-set-calibrated ANN model is verified. The early stopping method, however, needs one more subset called validation set between the training and the test sets. Whenever the mean-square error (MSE) of the validation set gets to its minimum amount, the training procedure would be terminated. The mean-square error begins to increase from its minimum amount if the ANN is trained further. Therefore, wherever the number of weights exceeds the sample size, the early stopping method can be used (Khademi et al. 2015, Jeong and Kim
3.1. Multiple Linear Regression Model (MLR) Generally, regression models can be considered as the process of fitting models to data. In order to handle the relationships between the variables, …show more content…
1989, Khademi et al. 2015). Hence, it is highly recommended to use a single hidden layer, and therefore, this recommendation has been used in this paper. Each particular node is a layer is connected to many other nodes using weighted connections. The weights estimate the strength of connections between interconnected nodes. The weights are trained in order to get the objective that the ANN`s output becomes similar to measured output on a recognized data …show more content…
The early stopping method has been known to be the most efficient method for optimizing the generalization performance of ANNs in practice. Therefore, early stopping method can be selected for both increasing the generalization capability of the model and overcoming the over-fitting. Normally, the data in ANN are divided into two sets of training and testing. Using the test subset, the performance of the training-set-calibrated ANN model is verified. The early stopping method, however, needs one more subset called validation set between the training and the test sets. Whenever the mean-square error (MSE) of the validation set gets to its minimum amount, the training procedure would be terminated. The mean-square error begins to increase from its minimum amount if the ANN is trained further. Therefore, wherever the number of weights exceeds the sample size, the early stopping method can be used (Khademi et al. 2015, Jeong and Kim