|
Title |
Comparison of Performance for Predicting Compressive Strength of Concrete Using Machine Learning
|
|
Authors |
이상민(Sang Min Lee) ; 성한석(Han Suk Sung) ; 강현구(Thomas H.-K. Kang) |
|
DOI |
https://doi.org/10.4334/JKCI.2022.34.5.505 |
|
Keywords |
기계학습; 압축강도; 배합설계; 신경망; 가우시안 프로세스 회귀모형 machine learning; compressive strength; mix design; neural networks; Gaussian process regression |
|
Abstract |
The purpose of this study is to compare and investigate the performance of various machine learning models for predicting the compressive strength of concrete. The compressive strength and mixing ratio data of the specimens measured from various experiments were combined with an open database. Input variables were related to cement, blast furnace slag, fly ash, water, superplasticizer, coarse aggregate, fine aggregate, and age. The output variable was the compressive strength of the concrete. As a consequence, the neural network and Gaussian process regression model outperformed the linear regression, decision tree, support vector machine, and ensemble models in terms of predictive performance. This study is superior in that it uses a machine learning model to predict the compressive strength of concrete within a limited error range without having to measure the compressive strength of concrete. This research paves the way to improving the accuracy of prediction through database accumulation and the incorporation of environmental variables in the future.
|