Nehal Elshaboury, Abobakr Al-Sakkaf, Ghasan Alfalah, Eslam Mohammed Abdelkader

Abstract:

Forecasting accurate labour productivity is critical in construction project management because construction projects are labour-intensive. This study proposes eight intelligent data-driven models for emulating formwork labour productivity in high rise buildings. These models encompass an adaptive neuro-fuzzy inference system trained using genetic algorithm (ANFIS-GA), an adaptive neuro-fuzzy inference system trained using particle swarm optimization algorithm (ANFIS-PSO), generalized regression neural network (GRNN), back-propagation artificial neural network (BP-ANN), Elman neural network (ENN), regression trees (RT), support vector machines (SVM) and Gaussian process regression (GPR). The models are applied to two high-rise buildings in Montreal, Canada to test their prediction capabilities. The accuracies of the developed data-driven models are investigated using the performance metrics of mean absolute percentage error (MAPE), mean absolute error (MAE), root-mean squared error (RMSE), root relative squared error (RRSE) and relative absolute error (RAE). The assessment metrics show that the GRNN model exhibits better and stable performance than the remainder of the prediction models (MAPE 8.98%, MAE 0.13, RMSE 0.19, RAE 0.45 and RRSE 0.54). It is also derived that the work method and temperature sustain the high influence on labor productivity. It can be anticipated that the developed GRNN model, can be a valuable decision-making tool for forecasting construction labour productivity in construction projects.

Keywords: Labour productivity, data-driven, high rise buildings, generalized regression neural network, regression tree, Gaussian process regression.

Date Published: January 6, 2022
DOI: 10.11159/ijci.2022.001

View Article