Scholarlink

A basic component of resources in civil engineering is concrete. It is applied in construction such as dams, bridges and buildings. Concrete compressive strength (CS) of the forms a huge aspect of the safety and durability. Admixtures are a proven means of improving concrete. Conventionally, evaluating compressive strength entails expensive and time-demanding laboratory assessments. To tackle this, machine learning frameworks like Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbour (KNN) were applied to forecast CS. The capacities of these models were assessed and measured using the Mean Absolute Error (MAE), Root-Mean-Squared Error (RMSE) and the Correlation Coefficient (R2). The RF framework performed better than other models with R2 score of 0.986, RMSE = 1.961 and MAE = 1.300. SVM recorded an RMSE score of 11.100, an R2 score of 0.567, MAE score of 8.146. KNN scored RMSE of 7.539, R2 of 0.800 and MAE of 5.612. Scatter plots, bar charts and other data visualization elements validate that RF model has proven to be correct concerning forecasting the CS of a concrete improved with admixture.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predicting Concrete Compressive Strength Using Various Machine-Learning Models

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Lanre Shittu,
  • Chukwuemeka Nwachukwu,
  • Chukwuebuka Akwiwu-Uzoma,
  • Samuel Ovuehor,
  • Kehinde Durodola-Tunde

摘要

A basic component of resources in civil engineering is concrete. It is applied in construction such as dams, bridges and buildings. Concrete compressive strength (CS) of the forms a huge aspect of the safety and durability. Admixtures are a proven means of improving concrete. Conventionally, evaluating compressive strength entails expensive and time-demanding laboratory assessments. To tackle this, machine learning frameworks like Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbour (KNN) were applied to forecast CS. The capacities of these models were assessed and measured using the Mean Absolute Error (MAE), Root-Mean-Squared Error (RMSE) and the Correlation Coefficient (R2). The RF framework performed better than other models with R2 score of 0.986, RMSE = 1.961 and MAE = 1.300. SVM recorded an RMSE score of 11.100, an R2 score of 0.567, MAE score of 8.146. KNN scored RMSE of 7.539, R2 of 0.800 and MAE of 5.612. Scatter plots, bar charts and other data visualization elements validate that RF model has proven to be correct concerning forecasting the CS of a concrete improved with admixture.