Abstract:This study explores the integration of machine learning (ML) techniques to predict and optimize the compressive strength of alkali-activated materials (AAMs) sourced from four industrial waste streams: blast furnace slag, fly ash, reducing slag, and waste glass. Aimed at mitigating the labor-intensive trial-and-error method in AAM formulation, ML models can predict the compressive strength and then streamline the mixture compositions. By leveraging a dataset of only 42 samples, the Random Forest (RF) model underwent fivefold cross-validation to ensure reliability. Despite challenges posed by the limited datasets, meticulous data processing steps facilitated the identification of pivotal features that influence compressive strength. Substantial enhancement in predicting compressive strength was achieved with the RF model, improving the model accuracy from 0.05 to 0.62. Experimental validation further confirmed the ML model's efficacy, as the formulations ultimately achieved the desired strength threshold, with a significant 59.65% improvement over the initial experiments. Additionally, the fact that the recommended formulations using ML methods only required about 5 min underscores the transformative potential of ML in reshaping AAM design paradigms and expediting the development process.

What problem does this paper attempt to address?

The paper attempts to address the problem of predicting and optimizing the compressive strength of alkali-activated materials (AAMs) prepared from four types of industrial waste (blast furnace slag, fly ash, reduction slag, and waste glass) using machine learning techniques. Specifically, the study aims to overcome the labor-intensive and time-consuming issues of traditional trial-and-error methods in AAM formulation design by predicting compressive strength with machine learning models and optimizing mixture components to improve the efficiency and performance of AAMs design. ### Main Objectives: 1. **Establish Machine Learning Models**: Utilize experimental data to establish machine learning models to accelerate the development of AAMs. 2. **Address Challenges of Small Datasets**: Overcome the challenges posed by small datasets in building machine learning models through data processing methods. 3. **Validate Model Effectiveness**: Validate the effectiveness of the established machine learning models by identifying AAM formulations with high compressive strength through experimental verification. ### Research Background: - **Alkali-Activated Materials (AAMs)**: AAMs are gaining attention for their environmental benefits and lower carbon footprint, utilizing industrial by-products and waste materials to further enhance their environmental benefits. - **Existing Challenges**: The application of AAMs faces engineering challenges such as workability loss, rheological issues, cost concerns, the need for special curing, and limited reactivity of precursors. - **Traditional Methods**: Traditional AAM design methods rely on trial-and-error experiments conducted by experienced operators, which are time-consuming and resource-intensive. - **Advantages of Machine Learning**: Machine learning algorithms can predict material performance and optimize mixture components, significantly reducing the need for physical experiments and accelerating the discovery process of new materials. ### Research Methods: 1. **Dataset**: The study used a dataset of 42 samples provided by the Materials and Chemical Research Laboratories (MCL) and the Industrial Technology Research Institute (ITRI). 2. **Data Processing**: This includes initial data processing, component transformation, component feature classification, feature selection, and feature construction. 3. **Model Evaluation**: Five-fold cross-validation and various performance metrics (such as Mean Absolute Error MAE, Root Mean Square Error RMSE, and Coefficient of Determination R²) were used to evaluate model performance. ### Experimental Results: - **Model Accuracy Improvement**: The accuracy of the Random Forest (RF) model improved from 0.05 to 0.62. - **Experimental Validation**: The recommended formulation achieved the required strength threshold in experiments, showing a 59.65% improvement over the initial experiment. - **Time Efficiency**: The formulation recommended by the machine learning method took only about 5 minutes, demonstrating the potential of machine learning in reshaping the AAM design paradigm and accelerating the development process. ### Conclusion: The study successfully applied machine learning techniques to predict and optimize the compressive strength of AAMs, overcoming the challenges posed by small datasets and validating the model's effectiveness through experiments. This provides important guidance and support for the future preparation of high-performance AAMs using industrial waste.

Efficient Compressive Strength Prediction of Alkali-Activated Waste Materials Using Machine Learning

Prediction of compressive strength and feature importance analysis of solid waste alkali-activated cementitious materials based on machine learning

Optimized Machine Learning Model for Predicting Compressive Strength of Alkali-Activated Concrete Through Multi-Faceted Comparative Analysis

Elemental Design of Alkali-Activated Materials with Solid Wastes Using Machine Learning

Application of Advanced Machine Learning Approaches to Predict the Compressive Strength of Concrete Containing Supplementary Cementitious Materials

Compressive strength prediction of one-part alkali activated material enabled by interpretable machine learning

Promoting low carbon construction using alkali-activated materials: A modeling study for strength prediction and feature interaction

Machine learning enabled closed‐form models to predict strength of alkali‐activated systems

Predictive Regression Models for the Compressive Strength of Fly Ash-based Alkali-Activated Cementitious Materials via Machine Learning

Machine Learning Modeling Integrating Experimental Analysis for Predicting Compressive Strength of Concrete Containing Different Industrial Byproducts

Prediction of compressive strength of alkali-activated construction demolition waste geopolymers using ensemble machine learning

Machine Learning Prediction Models to Evaluate the Strength of Recycled Aggregate Concrete

Accurate compressive strength prediction using machine learning algorithms and optimization techniques

Estimating the mechanical properties of high-performance concrete via automated and ensembled machine learning methods

Machine learning to predict properties of fresh and hardened alkali-activated concrete

Metaheuristic optimization of machine learning models for strength prediction of high-performance self-compacting alkali-activated slag concrete

Machine learning based modeling for predicting the compressive strength of solid waste material-incorporated Magnesium Phosphate Cement

Machine-Learning-Aided Prediction of Flexural Strength and ASR Expansion for Waste Glass Cementitious Composite

Predicting the compressive strength of steelmaking slag concrete with machine learning – Considerations on developing a mix design tool

Forecasting the Mechanical Properties of Plastic Concrete Employing Experimental Data Using Machine Learning Algorithms: DT, MLPNN, SVM, and RF

Predicting Compressive Strength of Concrete Containing Industrial Waste Materials: Novel and Hybrid Machine Learning Model