A Two-Level Machine Learning Prediction Approach for RAC Compressive Strength

Buildings, Год журнала: 2024, Номер 14(9), С. 2885 - 2885

Опубликована: Сен. 12, 2024

Through the use of recycled aggregates, construction industry can mitigate its environmental impact. A key consideration for concrete structural engineers when designing and constructing structures is compressive strength. This study aims to accurately forecast strength aggregate (RAC) using machine learning techniques. We propose a simplified approach that incorporates two-layer stacked ensemble model predict RAC In this framework, first layer consists models acting as base learners, while second utilizes random forest (RF) meta-learner. comparative analysis with four other demonstrates superior performance proposed in effectively integrating predictions from resulting enhanced accuracy. The achieves low mean absolute error (MAE) 2.599 MPa, root squared (RMSE) 3.645 high R-squared (R2) value 0.964. Additionally, Shapley (SHAP) additive explanation reveals influence interrelationships various input factors on RAC, aiding design professionals optimizing raw material content during production process.

Язык: Английский

---

Modeling properties of recycled aggregate concrete using gene expression programming and artificial neural network techniques

Frontiers in Built Environment, Год журнала: 2024, Номер 10

Опубликована: Окт. 10, 2024

Soft computing techniques have become popular for solving complex engineering problems and developing models evaluating structural material properties. There are limitations to the available methods, including semi-empirical equations, such as overestimating or underestimating outputs, and, more importantly, they do not provide predictive mathematical equations. Using gene expression programming (GEP) artificial neural networks (ANNs), this study proposes estimating recycled aggregate concrete (RAC) An experimental database compiled from parallel studies, a large amount of literature was used develop models. For compressive strength prediction, GEP yielded coefficient determination (R 2 ) value 0.95, while ANN achieved an R 0.93, demonstrating high reliability. The proposed both simple robust, enhancing accuracy RAC property estimation offering valuable tool sustainable construction.

Язык: Английский

---

Predictive methods for the evolution of oil well cement strength based on porosity

Materials and Structures, Год журнала: 2024, Номер 57(10)

Опубликована: Ноя. 4, 2024

Язык: Английский

---

Self-compacting concrete strength evaluation using fire hawk optimization-based simulations

Multiscale and Multidisciplinary Modeling Experiments and Design, Год журнала: 2024, Номер 8(1)

Опубликована: Ноя. 13, 2024

Язык: Английский

---

Self-Stressing State and Progressive Limit Method Study of a Flat Strip

Lecture notes in civil engineering, Год журнала: 2024, Номер unknown, С. 349 - 357

Опубликована: Дек. 31, 2024

Язык: Английский

---