Prediction of construction cost of high-rise residential buildings from owners’ perspective using a multiple-output neural network model | |
| Author | Gudipati, Pranathi |
| Call Number | AIT Thesis no.CM-23-08 |
| Subject(s) | Building--Estimates Construction projects--Estimates MATLAB Construction projects--Cost control--Mathematical models Neural networks (Computer science) |
| Note | A thesis submitted in partial fulfilment of the requirements for the degree of Master of Engineering in Construction, Engineering and Infrastructure Management |
| Publisher | Asian Institute of Technology |
| Abstract | The pre-design phase cost estimate is a critical aspect of any construction project, even though the project scope may not yet be finalized and there is limited information available during the early stages. This study aimed to identify the most significant parameters affecting the construction cost and develop an efficient model for predicting the cost of building construction projects using artificial neural networks. Data from 74 high-rise residential buildings were collected from the construction industry in Hyderabad, a city in south India, and several key parameters were identified for the construction cost of residential buildings. The developed artificial neural network (ANN) model included an input layer with 15 neurons, one hidden layer with 15 neurons, and an output layer with 7 neurons representing excavation cost, foundation cost, basement cost, super-structure cost, finishings cost, labour cost, and total cost of the building per square meter. The results indicated that neural networks were successful in predicting early-stage cost estimates of buildings using basic information, without the need for more detailed design with 94.52% accuracy. The predictor importance analysis showed that several factors, including type of soil, formwork used in construction, slab area of the building, topography of the land before construction and number of units, were the most influential parameters affecting the prediction of construction costs. Interestingly, the formwork used in construction was found to be one of the most influential parameters, a new finding that adds to the existing body of knowledge in this area. Overall, the study underscores the importance of accurate cost estimates during the pre-design phase of construction projects, and highlights the potential of artificial neural networks in improving these predictions. |
| Year | 2023 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Construction Engineering and Infrastructure Management (CM) |
| Chairperson(s) | Hadikusumo, Bonaventura H. W. |
| Examination Committee(s) | Tripathi, Nitin Kumar;Shanmugam, Mohana Sundaram |
| Scholarship Donor(s) | AIT Scholarships |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2023 |