Sensitivity Analysis of Risk Management In Construction Projects Using Deep Learning Techniques
The construction project plays an important role in the economy and welfare of the people of every society. Every year, large budgets are spent on the implementation of construction projects in Iran. Risk management is the missing link in the studies and implementation of construction projects in Iran, which is either less paid or is not a priority. In this article, an attempt has been made to deal with the phenomenon of risk management of construction projects. First, the risk management of the construction project is done using the FMEA method, then modeling this phenomenon using the deep learning method is on the agenda. The database for deep learning modeling has been completed using the information of the first stage. The input and output parameters are also selected based on the number and type of available data. The research results show that the optimal structure of the deep learning method has been introduced by combining the CNN neural network and the MVO optimization algorithm, with a correlation coefficient above 90% and an error index below 0.5. The main discussion of this research is sensitivity analysis on the optimal model. This process has been done by using the relative and absolute derivative of the output with respect to each input based on different statistical categories according to the method of Lu et al. Finally, the most effective input parameter is introduced and the output sensitivity is determined.