Application of knowledge discovery in database (KDD) techniques in cost overrun of construction projects

Currently, cost overrun is a global challenge to completing construction projects successfully. To overcome this problem, earlier studies investigated factors of cost overrun. Knowledge Discovery in Data (KDD) and data mining techniques have been implemented effectively in various research areas to extract novel and valuable knowledge from historical data but have only recently been implemented in the construction industry. The aim of this research is to develop a model that predicts project cost overrun using a suitable data mining technique and cost overrun factors as predictors. A review of the literature identified twelve factors that can be easily measured and analyzed in construction projects. A case study was performed to validate the model with an actual data set of executed projects. The resulting model is simple, interpretable, and relatively accurate (60.87%), and it uses three steps of data mining–clustering, feature selection, and classification. These steps improve model performance.