Hybrid-based Deep Belief Network Model for Cement Compressive Strength Prediction

Compressive strength is one of the most important qualities of concrete, and most of the conventional regression models for predicting the concrete strength could not achieve an expected result due to the unstructured factors. Moreover, the utilization of machine learning and statistical approaches playing its vital role in predicting the concrete compressive strength based on mixture proportions accounting to its industrial importance as well. In this manner, this paper attempts to introduce a new deep learning-based prediction model that makes the prediction more accurate, hence Deep Belief Network (DBN) is used. Moreover, to make the prediction more precise, it is planned to have the fine-tuning of activation function and weights of DBN, which makes the model efficient in its performance. For this purpose, an improved optimization concept is introduced called Lion Algorithm with new Rate Evaluation, which is the modified Lion Algorithm (LA). Finally, the performance of the proposed model is evaluated over other state-of-the-art models concerning certain error analysis.