Predicting financial distress using hybrid feedforward neural network with cuckoo search algorithm

Abstract Research on predicting financial distress is still an ongoing topic. Artificial neural networks (ANNs) models are good candidates to model financial distress due to their abilities for detecting complex patterns. In this context, this study looked at the performance of advanced cuckoo search algorithm to seek optimum weights of the feedforward neural network(FNN). This model called CSFNN was compared with Logistic regression (LR) and backpropagation feedforward neural network (BPNN) in order to investigate its efficiency. We used data collected in manufacturing sector at two different periods, one year and three years before the bankruptcy. We found that the CSFNN model led to 90.30% accuracy whereas the BPNN and LR models led to 88.33% and 82.15% accuracy respectively at one year before the bankruptcy. For three years before the bankruptcy, we find that the CSFNN performed at 82.79% followed by the BPNN at 81.05% and the LR at 73.27%. In conclusion, experimental results shown that the cuckoo search algorithm in a FNN model could be considered for predicting potential financial distress.

[1]  Haibo He,et al.  Learning from Imbalanced Data , 2009, IEEE Transactions on Knowledge and Data Engineering.

[2]  Gustavo E. A. P. A. Batista,et al.  A study of the behavior of several methods for balancing machine learning training data , 2004, SKDD.

[3]  Vikas Kumar,et al.  Systematic review of bankruptcy prediction models: Towards a framework for tool selection , 2018, Expert Syst. Appl..

[4]  J. March.  Exploration and exploitation in organizational learning , 1991, STUDI ORGANIZZATIVI.

[5]  Qinghua Huang,et al.  Predicting financial distress and corporate failure: A review from the state-of-the-art definitions, modeling, sampling, and featuring approaches , 2014, Knowl. Based Syst..

[6]  Huimin Lu,et al.  Brain Intelligence: Go beyond Artificial Intelligence , 2017, Mobile Networks and Applications.

[7]  Antanas Verikas,et al.  Hybrid and ensemble-based soft computing techniques in bankruptcy prediction: a survey , 2010, Soft Comput..

[8]  Jakub M. Tomczak,et al.  Ensemble boosted trees with synthetic features generation in application to bankruptcy prediction , 2016, Expert Syst. Appl..

[9]  Michal Tkác,et al.  Artificial neural networks in business: Two decades of research , 2016, Appl. Soft Comput..

[10]  Herbert Kimura,et al.  Machine learning models and bankruptcy prediction , 2017, Expert Syst. Appl..

[11]  Ning Chen,et al.  Financial credit risk assessment: a recent review , 2015, Artificial Intelligence Review.

[12]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[13]  Aníbal R. Figueiras-Vidal,et al.  Pattern classification with missing data: a review , 2010, Neural Computing and Applications.

[14]  Saeed Tavakoli,et al.  Improved Cuckoo Search Algorithm for Global Optimization , 2011 .

[15]  Xin-She Yang,et al.  Cuckoo search: recent advances and applications , 2013, Neural Computing and Applications.

[16]  Philippe du Jardin,et al.  Predicting bankruptcy using neural networks and other classification methods: The influence of variable selection techniques on model accuracy , 2010, Neurocomputing.

[17]  Václav Snásel,et al.  Metaheuristic design of feedforward neural networks: A review of two decades of research , 2017, Eng. Appl. Artif. Intell..