Predicting compressive strength of concrete using machine learning

Abstract Concrete compressive strength is verified by destructive testing after curing, which limits how quickly candidate mixtures can be screened. This study evaluates whether mixture proportions and curing age support accurate, interpretable preliminary estimates, and how much the reported accuracy depends on the evaluation protocol. Linear Regression, Support Vector Regression, Random Forest, and Gradient Boosting were compared on the UCI Concrete Compressive Strength dataset (1,030 observations, 428 mixtures). A strength-stratified 80/20 split gave 824 training and 206 held-out observations; hyperparameters were selected by five-fold cross-validation within the training partition and the held-out set was used once. Gradient Boosting performed best ( $$R^2=0.940$$ , RMSE = 4.07​ MPa, MAE = 2.70 MPa), followed by Random Forest ( $$R^2=0.926$$ ) and Support Vector Regression ( $$R^2=0.887$$ ). Because most mixtures are tested at several ages, random splits estimate generalisation to new observations of known mixtures; mixture-grouped cross-validation, which targets new mixtures, lowered the Gradient Boosting score from 0.933 to 0.862 with the ranking preserved. SHAP attributions ranked age, cement, water, and slag as the leading predictors, consistent with established concrete technology but not read as causal evidence. The framework supports preliminary mixture screening with laboratory verification; external validation remains necessary.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-13
DOI
https://doi.org/10.1038/s41598-026-79849-5
Primary Topic
Concrete and Cement Materials Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Predicting compressive strength of concrete using machine learning

Samer El-Zahab, Hassan Ghanem, Kotaiba Mokadam, Furkan Uysal et al.
1 links
Scientific Reports
Concrete and Cement Materials Research
article

Predicting compressive strength of concrete using machine learning

Samer El-Zahab, Hassan Ghanem, Kotaiba Mokadam, Furkan Uysal, Shareefa Baqer, Retaj Aldehani, Fouz Alhaji, Malak Alzaidi, Amirah Rasheed
article en

Abstract

Abstract Concrete compressive strength is verified by destructive testing after curing, which limits how quickly candidate mixtures can be screened. This study evaluates whether mixture proportions and curing age support accurate, interpretable preliminary estimates, and how much the reported accuracy depends on the evaluation protocol. Linear Regression, Support Vector Regression, Random Forest, and Gradient Boosting were compared on the UCI Concrete Compressive Strength dataset (1,030 observations, 428 mixtures). A strength-stratified 80/20 split gave 824 training and 206 held-out observations; hyperparameters were selected by five-fold cross-validation within the training partition and the held-out set was used once. Gradient Boosting performed best ( $$R^2=0.940$$ , RMSE = 4.07​ MPa, MAE = 2.70 MPa), followed by Random Forest ( $$R^2=0.926$$ ) and Support Vector Regression ( $$R^2=0.887$$ ). Because most mixtures are tested at several ages, random splits estimate generalisation to new observations of known mixtures; mixture-grouped cross-validation, which targets new mixtures, lowered the Gradient Boosting score from 0.933 to 0.862 with the ranking preserved. SHAP attributions ranked age, cement, water, and slag as the leading predictors, consistent with established concrete technology but not read as causal evidence. The framework supports preliminary mixture screening with laboratory verification; external validation remains necessary.

Scientific Reports
American University of the Middle East (KW), Beirut Arab University (LB)
Concrete and Cement Materials Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon