Prediction of long-term deflection of reinforced concrete beams using Jellyfish Search optimization machine learning model

Abstract

The long-term deflection of reinforced concrete beams has always been a challenge in structural design. This study focuses on developing a machine learning model using the Jellyfish Search optimization algorithm to predict the long-term deflection of reinforced concrete beams. Based on a dataset from published research, machine learning models were built and evaluated (including single and ensemble models) to select the most accurate model. The Jellyfish Search optimization algorithm was used to optimize the parameters of the selected model. The comparison results showed that the JS – Bagging ANN model achieved superior performance with R = 0.976; MAE = 3.988 mm; RMSE = 1.777 mm; MAPE = 14.154%; and SI = 0.00 (1). Therefore, the JS - Bagging ANN model is highly recommended for predicting the long-term deflection of reinforced concrete beams in structural design calculations.

Keywords: Long - term deflection; reinforced concrete beam; machine learning model; jellyfish search; optimization.