Abstract—The bankruptcy prediction model (BPM) is eminently essential for financial institutions to verify the creditworthiness of companies or management. The inability to accurately predict the bankruptcy can destroy the effects of socio-economics. Hence, it is significant to offer financial decision makers with efficient bankruptcy prediction to forestall these loss states. This paper presents a comprehensive review based on various statistical and machine learning techniques to address the issue of bankruptcy prediction. The statistical techniques include, linear discriminant analysis (LDA), multivariate discriminant analysis (MDA) and logistic regression (LR), and machine learning techniques such as artificial neural networks (ANN), support vector machines (SVM) and decision trees. Traditional statistical techniques were used to handle large data sets without affecting the prediction performance. Furthermore, the machine learning techniques provide greater prediction accuracy than the traditional statistical techniques for smaller data sets. Besides, optimization techniques, such as genetic algorithm (GA) and particle swarm optimization (PSO), were integrated with machine learning techniques to further improve the prediction accuracy for large data sets. This paper conducts a comparative analysis of the various techniques used based on their corresponding benefits and limitations. In our future work, the prediction of bankruptcy may be improved by integrating other heuristic evolutionary algorithms with the machine learning techniques using the Apache Mahout tool.
Index Terms—Artificial neural networks (ANN), bankruptcy prediction model (BPM), optimization techniques and support vector machines (SVM).
The authors are with the Dept. of CSE, GIT, GITAM University, Visakhapatnam, Andhra Pradesh, India (e-mail: email@example.com, firstname.lastname@example.org).
Cite: S. Sarojini Devi and Y. Radhika, "A Survey on Machine Learning and Statistical Techniques in Bankruptcy Prediction," International Journal of Machine Learning and Computing vol. 8, no. 2, pp. 133-139, 2018.