VISUAL MATERIAL SEGREGATION JUDGMENT PREDICTION OF HIGH-FLUID CONCRETE BY USING MACHINE LEARNING

Abstract

In this study, we attempted to use Machine Learning to predict the visual judgment of material segregation of High-fluid concrete using the materials and concrete formulation as features. Random Forest and LightGBM, a type of ensemble learning, were used to obtain feature importance. Furthermore, Logistic Regression, a type of statistics and Machine Learning, was performed using these features of high importance in Machine Learning. Using Logistic Regression, we proposed a prediction formular for visual material segregation judgment and the prediction performance of the proposed prediction formula was evaluated.