Advanced prediction of compressive strength of Ceramic Tile Waste (CTW) modified concrete using Gene Expression Programming (GEP)
AI-Powered Analysis: Predicting Sustainable Concrete Strength
Leveraging Gene Expression Programming for advanced material science to optimize concrete mix designs.
Executive Summary & Key Impact
This study applies Gene Expression Programming (GEP) to predict the compressive strength of ceramic tile waste-modified concrete, addressing environmental concerns of natural sand overexploitation. A dataset of 136 unique concrete mixes from published literature was used. Input variables include cement, fine and coarse natural aggregates, ceramic fine aggregate, curing time, superplasticizer, water, and water-cement ratio. The GEP10 model achieved the best performance (R=0.916 training, R=0.933 validation), demonstrating strong predictive accuracy with RMSE of 4.723 MPa and MAE of 3.713 MPa. SHAP analysis revealed curing age and water content as the most influential features. The results highlight GEP's robustness for sustainable concrete modeling.
0.933
Validation R²
4.723 MPa
RMSE (MPa)
3.713 MPa
MAE (MPa)
Age
Key Influencer
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
The study employs Gene Expression Programming (GEP) to predict concrete compressive strength. It involves literature review, data collection, preprocessing, GEP model development, optimization, and validation. The dataset, compiled from 136 mixes in published literature, includes eight input variables (cement, aggregates, curing time, superplasticizer, water, water-cement ratio) and one output variable (compressive strength). Data was split 70/30 for training/validation. Model performance was evaluated using MAE, RMSE, R, and RRMSE. SHAP analysis was used for interpretability.
GEP Methodology Flowchart
Literature Review
→
Data Collection (from peer review sources)
→
Data Preprocessing
→
Data cleaning
→
Statistical measures and visualization
→
Features Scaling
→
Machine Learning Models
→
Model Development
→
Model Optimization
→
Model Validation
136
Concrete Mixes Analyzed
GEP Model 10 emerged as the best performer, with a validation R of 0.933 and RMSE of 4.723 MPa. SHAP analysis identified curing age and water content as the most influential features, collectively contributing 55% to predictive capability. Aggregates and admixtures had minimal impact. The study emphasizes GEP's ability to model nonlinear behavior and provide a reliable, interpretable framework for sustainable concrete.
| Model | Training R | Validation R | Validation RMSE (MPa) |
|---|---|---|---|
| GEP10 (Optimal) | 0.916 | 0.933 | 4.723 |
| GEP20 | 0.912 | 0.926 | 5.105 |
| GEP03 | 0.919 | 0.897 | 5.835 |
| GEP19 | 0.91 | 0.91 | 5.643 |
Impact of Influential Features
SHAP analysis revealed that Age and Water content are the most influential features, contributing 37% and 18% respectively to the model's predictive capability. Higher curing periods positively correlate with strength, while excessive water content negatively impacts it due to increased porosity. Conversely, aggregates (coarse, fine, ceramic) and superplasticizer showed relatively negligible influence on predictions. This insight is crucial for optimizing concrete mix designs for desired compressive strengths.
37%
Age Contribution
18%
Water Content Contribution
The GEP model offers a cost-effective and time-saving alternative to traditional laboratory testing for ceramic tile waste-modified concrete. Its ability to provide interpretable mathematical equations fosters transparency, reducing the 'black box' problem often associated with other ML models. This predictive framework supports sustainable construction by optimizing resource utilization and minimizing waste, accelerating the adoption of eco-friendly building materials.
Reduced Time & Cost
Traditional vs. AI Testing
Advanced ROI Calculator: Optimize Your Concrete Mix Design
Estimate the potential cost savings and efficiency gains for your enterprise by implementing AI-driven concrete mix prediction. Adjust the parameters below to see the impact on your operations.
Annual Savings
$0
Hours Reclaimed Annually
0
Your AI Implementation Roadmap for Sustainable Concrete
A structured approach to integrating AI-driven concrete mix design into your enterprise operations.
Phase 1: Data Assessment & Integration
Review existing concrete mix data, integrate with GEP model, and establish data pipelines for ceramic tile waste parameters.
Phase 2: Model Customization & Training
Tailor the GEP model to your specific material sources and project requirements, refining parameters with your internal datasets.
Phase 3: Pilot Project & Validation
Apply the GEP-predicted mixes to a pilot construction project, validating actual compressive strengths against predictions.
Phase 4: Full-Scale Deployment & Monitoring
Integrate the GEP solution across all relevant projects, continuously monitoring performance and refining the model over time.
Schedule Your AI Strategy Session
Unlock the full potential of AI in sustainable concrete design. Book a personalized consultation with our experts to discuss how GEP can revolutionize your material science and construction projects.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat