# Explainable machine learning model and gene expression programming for predicting reinforced concrete beams moment capacity exposed to fire

## TL;DR

This study uses machine learning and gene expression programming to predict the strength of reinforced concrete beams after exposure to fire.

## Contribution

A new formulation for predicting moment capacity of RC beams under fire using GEP and ML models is proposed.

## Key findings

- XGBoost model achieved the highest accuracy with the best R2 and lowest error rate.
- SHAP analysis provided interpretable insights into the XGBoost model's predictions.
- The proposed methods demonstrated high accuracy in estimating moment capacity of RC beams under fire conditions.

## Abstract

In this study, a new formulation for the moment capacity (Mr) of Reinforced Concrete (RC) beams under fire conditions is estimated using Gene Expression Programming (GEP). In addition, the use of Machine Learning (ML) methods such as XGBoost, AdaBoost, and LightGBM is investigated for estimating the Mr of RC beams in fire. The database for predicting the Mr of RC beams includes 280 samples. In this paper, the cross-section width bw, cross-section depth d, distance from the beam edge to the center of steel reinforcement deff, area of steel reinforcement Ast, time duration of fire t, compressive strength of concrete fc, and moment capacity of the beam under fire Mr are considered as the parameters of ML models. Several statistical metrics were employed to assess the performance of the models, including the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), coefficient of determination (R2), and gradients of regression lines (k and k′). In this study, Shapley Additive exPlanations (SHAP) analysis was used to interpret the predictions of the XGBoost model, which was selected for its high accuracy with the best R2 and the lowest error rate. The results indicate that the methods used demonstrate high accuracy in estimating the Mr of RC beams.