Using artificial neural network (ANN) to estimate the cost of residential building project in the feasibility phase

Cost is considered as one of the main challenges for each project manager in the construction industry and one of the criteria that measure the success of the project is the comparison between the cost of the project with the planned budget at the beginning of the project. For this purpose, an accurate forecast in the early stages of the project can greatly assist the decisions that are taken in the feasibility phase. This accurate cost estimate will help us choose the best implementation systems in different parts of the building according to the budget we plan, and avoid the options that may not lead to our main objectives in the project. in this study, we intend to design a model to predict the cost of construction in a feasibility stage using artificial neural network (ANN) as one of the tools of artificial intelligence field, whose error is less than traditional methods and calculation time is much faster than methods such as bottom-top cost estimation. for this purpose, we have identified the most important factors that affect construction cost by experts in housing investment group and then we have collected data from the past projects. In the next stage we have used MATLAB program to build the construction cost estimation model using ANN based on data gathered from past projects on housing investment group. In the end, the model based on the obtained results has a precision of 89.56 %. Finally, a graphic user interface is designed as an EXE. file, which facilitates the use of the model for users to estimate the cost of new projects and eliminates the need for a MATLAB application to use the model.