Prediction of Flow Properties of High-fluidity Concrete Using Multiple Machine Learning Models
Year: 2025
Type: article
Abstract: This study extended previous research by predicting the slump flow and time to 500 mm flow using various factors of high-fluidity concrete with multiple machine learners. The prediction results for unknown data showed a coefficient of determination R2 of 0.73 for slump flow and 0.71 for 500 mm flow arrival time, both with high accuracy over 70%. Permutation Feature Importance and Partial Dependence Plot visualized and evaluated the feature importance and influence of various factors on flow characteristics through machine learning.
DOI: 10.3130/aijt.31.625
Authors: Retsu MIURA, Kouta NAMIHIRA, Yoshitomo YAMADA, Syuya HIRANO
Institution: University of the Ryukyus
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4411443731