University of Twente Student Theses

Login

Predicting construction costs in the program phase of the construction process: a machine learning approach

Beltman, J.F. (2021) Predicting construction costs in the program phase of the construction process: a machine learning approach.

[img] PDF
1MB
Abstract:Cost estimation in the program phase of the construction process can only be roughly estimated at the moment of writing. On the other hand, the use of data will become more important in the future and with sufficient data, statistical predictions can be made using machine learning. Following this trend, machine learning could possibly improve the cost estimation in the program phase of the construction process. The conventional cost estimation method relies on a combination of analogous estimation and expert judgement. Machine learning, which is classified as parametric cost estimation, could initiate a shift in cost estimation methods. Earlier studies that used machine learning to perform cost estimation in the construction sector yielded promising results. The objective for a solution in this thesis is defined as when the created machine learning model could generate predictions within a 20% error range. Several building quantities have been used as input for the machine learning model. The machine learning types that have been selected are random forest and support vector machine. Two classes of building types have been used to train the model two separate times. Respectively 48% and 46% of all predictions of the two project classes were within a 20% error range. This contradicted the expectations and earlier studies. Evaluation of the model turned out that the model could potentially not predict the cost specific enough.
Item Type:Essay (Bachelor)
Clients:
Arcadis, Arnhem, Netherlands
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:01 general works
Programme:Industrial Engineering and Management BSc (56994)
Link to this item:http://purl.utwente.nl/essays/88257
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page