Abstract
The prediction of failure of a firm is a challenging topic in business research. In this paper, we consider a machine learning approach to detect the state of asset shortfall in the Italian small and medium-sized enterprises’ context. More precisely, we use the recurrent neural networks to predict the insolvency of firms. The huge dataset we study allows us to overcome problems of distortions given by smaller sample sizes. The observed sample comes from AIDA database, and consider thirty variables replicated for five years. The main result is that recurrent neural networks outperform the multi-layer perceptron architecture used as benchmark. The obtained accuracy scores are in line with those found in the literature, and this suggests that the use of new techniques such as those tried out in this study could produce even better results.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
procedure/cessazione indicates that a pending procedures concerning the change of legal status for the considered firm is in place.
- 2.
Generally speaking, Cohen’s kappa and \(F_1\) score are both known measures of classification accuracy. The former indicates substantial agreement between the considered model and the investigated phenomenon for values greater than 0.80 [9], closer the latter to the value 1 higher is the accuracy [16].
- 3.
Type I error is the share of firms incorrectly classified among all the firms that failed, while Type II error is the number of healthy firms that were misclassified (based on the networks trained in Phase 1).
References
Altman E.I., Sabato G.: Modelling Credit Risk for SMEs: Evidence from the U.S. Market. Abacus 43(3) 332–357 (2007)
Ba J.L., Kingma D.P.: ADAM: a method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2015)
Brédart, X.: Bankruptcy prediction model using neural networks. Account. Financ. Res. 3(2), 124–128 (2014)
Boritz, J.E., Kennedy, D.B.: Effectiveness of neural network types for prediction of business failure. Expert. Syst. Appl. 9(4), 503–512 (1995)
Glorot, X., Bordes, A., Bengio, Y.: Deep sparse rectifier neural networks. In: Proceedings of the 14th International Conference on Artificial Intelligence and Statistics (AISTATS) (2011)
Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning, 2nd edn. Springer (2009)
Hochreiter, S., Schmidhuber, J.: Long short term memory. Neural Comput. 9(8), 1735–1780 (1997)
Kumar, P.R., Ravi, V.: Bankruptcy prediction in banks and firms via statistical and intelligent techniques - A review. Eur. J. Oper. Res. 180(1), 1–28 (2007)
Landis, J.R., Koch, G.G.: The measurement of observer agreement for categorical data. Biometrics 33(1), 159–174 (1977)
LeCun, Y., Bottou, L., Orr, G.B., Müller, K.-R.: Efficient BackProp. In: Montavon, G., Orr, G.B., Müller, K.-R. (eds.) Neural networks: Tricks of the trade, pp. 9–50. Springer (1998)
Le, H.H., Viviani, J.L.: Predicting bank failure: an improvement by implementing a machine-learning approach to classical financial ratios, Research in International Business and Finance, 44(C), 16–25 (2018)
Lipton, Z.C., Berkowitz, J., Elkan, C.: A critical review of recurrent neural networks for sequence learning, arXiv preprint arXiv:1506.00019 (2015)
Louzada, F., Ara, A., Fernandes, G.B.: Classification methods applied to credit scoring: systematic review and overall comparison. Surv. Oper. Res. Manag. Sci. 21(2), 117–134 (2016)
Mitchell, T.M.: Machine Learning. McGraw-Hill (1997)
Ng A.Y: Feature selection, L1 versus L2 regularization, and rotational invariance. In: Proceedings of the 21st International conference on Machine Learning (2004)
Powers, D.M.W.: Evaluation: from precision, recall and F-measure to ROC, informedness, markedness & correlation. J. Mach. Learn. Technol. 2(1), 37–63 (2011)
Rosenblatt F.: The Perceptron: a perceiving and recognizing automaton, Report 85-460-1, Cornell Aeronautical Laboratory Inc. (1957)
Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent Neural Networks from overfitting. J. Mach. Learn. Res., 15(Jun), 1929–1958 (2014)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Nadali, L., Corazza, M., Parpinel, F., Pizzi, C. (2020). Recurrent ANNs for Failure Predictions on Large Datasets of Italian SMEs. In: Esposito, A., Faundez-Zanuy, M., Morabito, F., Pasero, E. (eds) Neural Approaches to Dynamics of Signal Exchanges. Smart Innovation, Systems and Technologies, vol 151. Springer, Singapore. https://doi.org/10.1007/978-981-13-8950-4_14
Download citation
DOI: https://doi.org/10.1007/978-981-13-8950-4_14
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-13-8949-8
Online ISBN: 978-981-13-8950-4
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)