Table 1 Benefits of the ANN, AdaBoost and SVM over conventional methods31,27,33.

From: Predicting the compressive strength of polymer-infused bricks: A machine learning approach with SHAP interpretability

Features

Adopted methods

Conventional methods

Handling Complexity

Effectively manages high-dimensional spaces and complex data

May struggle with high-dimensional data and complex relationships

Adaptability

Highly adaptable; can model complex, nonlinear relationships without predefined assumptions

Limited adaptability; often based on fixed assumptions and empirical formulas

Model Transparency

Enhanced transparency with techniques like SHAP for interpreting model predictions

Often lacks transparency; models are not always explainable

Accuracy

Often achieves higher accuracy due to advanced algorithms that capture intricate patterns and interactions

Accuracy depends on the quality and quantity of data; may require extensive empirical adjustments