Profitability Prediction for Construction Projects by Neural Network with Autoregressive Model: A Detail Analysis

This study examines the variables influencing the calculation of the profit ratio throughout the building project from the tendering process to completion. To identify the key elements controlling the calculation of profit ratio, the impact of projects on aspects by owner, engineer, contractor, clients, manufacturer and labourers was examined. Delay fines and client experience were the least significant in determining the profit ratio. The net profit was decided by the primary determinants of a company. Two distinct net profit models were created: the Autoregressive Integrated Moving Average Model (ARIMA) and the Artificial Neural Networks (ANN)-based dispute attribute prediction model. The objective of this optimisation issue is to minimise the aforementioned error between the target and output by determining the weights and biases of the ANN. The ANN methodology was used to ensure that these elements are the main factors influencing project disputes and to create a dispute prediction model for them. In this regression analysis, the factors influencing the profitability of construction are analysed clearly for better understanding. With inputs that reflect factors that were ranked as the most influential, an ANN was created to forecast the project performance model. The proposed models reduce the degree of construction risk in the profit ratio and enable project managers to concentrate on the critical success criteria.