Predicting Business Failure of Construction Contractors Using Long Short-Term Memory Recurrent Neural Network
- Authors
- Jang, Youjin; Jeong, In-Bae; Cho, Yong K.; Ahn, Yonghan
- Issue Date
- Nov-2019
- Publisher
- American Society of Civil Engineers
- Keywords
- Business failure; Construction contractors; Prediction model; Long short-term memory (LSTM); Recurrent neural network (RNN)
- Citation
- Journal of Construction Engineering and Management - ASCE, v.145, no.11, pp 1 - 9
- Pages
- 9
- Indexed
- SCIE
SCOPUS
- Journal Title
- Journal of Construction Engineering and Management - ASCE
- Volume
- 145
- Number
- 11
- Start Page
- 1
- End Page
- 9
- URI
- https://scholarworks.bwise.kr/erica/handle/2021.sw.erica/2244
- DOI
- 10.1061/(ASCE)CO.1943-7862.0001709
- ISSN
- 0733-9364
1943-7862
- Abstract
- Predicting business failure of construction contractors is critical for both contractors and other stakeholders such as project owners, surety underwriters, investors, and government entities. To identify a new model with better prediction of business failure of the construction contractors, this study utilized long short-term memory (LSTM) recurrent neural network (RNN). The financial ratios of the construction contractors in the United States were collected, and synthetic minority oversampling technique (SMOTE) and Tomek links were employed to obtain a balanced data set. The proposed LSTM RNN model was evaluated by comparing its accuracy and F1-score with feedforward neural network (FNN) and support vector machine (SVM) models for the optimized parameters selected from a grid search with five-fold cross-validation. The results successfully demonstrate that the prediction performance of the proposed LSTM RNN model outperforms FNN and SVM models for both test and original data set. Therefore, the proposed LSTM RNN model is a promising alternative to assist managers, investors, auditors, and government entities in predicting business failure of construction contractors, and can also be adapted to other industry cases.
- Files in This Item
-
Go to Link
- Appears in
Collections - COLLEGE OF ENGINEERING SCIENCES > MAJOR IN ARCHITECTURAL ENGINEERING > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.