An Accurate Deep Neural Network Model for Shear Strength of Light Weight Concrete Elements Based on COVID-19 Optimization

Concrete element shear strength is a conundrum that is controlled by a variety of factors and procedures. The type of concrete is thought to have the greatest influence since it alters the material's characteristics and behavior. Therefore, it requires the utmost attention. Because it has better qualities than regular eight concrete, lightweight concrete is one of the most preferred kinds of concrete. Based on the COVID-19 optimization, an effective deep neural network (DNN) model is created in this work to predict the shear strength of lightweight concrete components. Choosing the appropriate starting weights and bias of the DNN model is the function of the COVID-19 optimization. The proposed DNN model was validated by comparing it with other international design codes (ACI, EC2 & JSCE) to evaluate the success of this research in terms of enhancing the prediction accuracy. The comparing results show that the DNN model based on the COVID-19 optimization gives more accurate values and minimum error with respect to other traditional models.