Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
An Artificial Intelligence Framework to Contractor Financial Prequalification
Financial distress in the construction industry always causes major disruptions that usually result in a rippling effect on the economy. Avoiding such defaults is a top priority for employers to meet their demands. Artificial Intelligence (AI) models have provided increased accuracy in predicting fi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
AUC Knowledge Fountain
2023
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | Financial distress in the construction industry always causes major disruptions that
usually result in a rippling effect on the economy. Avoiding such defaults is a top priority for employers to meet their demands. Artificial Intelligence (AI) models have provided increased accuracy in predicting financial distress compared to statistical, fuzzy and logistic regression models, and other classification models. The main objective of this work is to support project employers in pre-qualifying contractors by predicting the status of construction contractors during a bid analysis to disqualify contractors with a high probability of experiencing financial distress during the project duration. Eight financial indicators & six macroeconomic variables were used in the analysis. The selected variables were proven to be highly correlated with the output values as provided in the literature while maintaining variables with diverse effects on the output. This work employs multiple models including artificial neural networks (ANN), support vector machines (SVM), and logistic regression using different tools (Python & NeuralTools) based on collected financial statements and macroeconomic indicators. The results show that the ANN model developed using python achieved higher performance measures than SVM (radial basis function & linear kernel functions), logistic regression & ANN developed using NeuralTools. The results also show that adding macroeconomic variables to financial ratios as input variables significantly enhance the accuracy and F-1 score of the model. Accordingly, the developed model is effective in predicting financial distress for construction companies. |
|---|