Prediction of Cost and Delay of Construction Projects Using Artificial Intelligence Techniques

Abstract

abstract The construction industry is subject to a high level of risks and uncertainties than any other industry. In reality, most participants experience risks in cost and time overruns and often fail to meet quality standards and operational requirements. In order to overcome these risks and make decisions with high accuracy, traditional and smart techniques have been applied to predict the cost and delay of construction projects with a high degree of accuracy and minimal errors. This research aimed to investigate the accuracy of five Artificial Intelligent techniques (Artificial Neural Network, Support Vector Machine, Extreme Learning Machines, Artificial Neural Network-Particle Swarm Optimization, and the Adaptive Neuro- Fuzzy Inference System) to demonstrate the impact of risk factors on prediction the cost and delay of construction projects. These techniques were represented through a virtual graphical user interface allowing the user to ease and clarify use. This study collected data from 47 construction projects from the AL-ZAWRAA state company in Baghdad city. Thus, the factors of risk were specified as well as analyzed employed Probability and Impact Analysis which were adopted as the inputs of the models. In contrast, the outputs of models were represented ratio to the costs of project and the delay in the construction project. viii Root Mean Squared Error (RMSE), correlation coefficient (R), and the coefficient of determination (R2) were utilized as the indices of performance of the models to evaluate the results accuracy. The results showed that the optimal method based on root mean squared error, and enabling to predict the cost and delay of projects was (ELM) with percentage (0.003) while the optimal method based on correlation factor and coefficient of determination were (ANFIS and ELM) with percentage (0.999, 0.999) and (0.999, 0.999) respectively. It was concluded that artificial intelligence techniques could be used as successful tools to solve essential problems in construction projects, especially in estimating the costs and delays. Besides, it supported construction companies in analyzing and evaluating risks affecting the management of new projects.