Bankruptcy prediction model using cost-sensitive extreme gradient boosting in the context of imbalanced datasets

Abstract

<span lang="EN-US">In the process of bankruptcy prediction models, a class imbalanced problem has occurred which limits the performance of the models. Most prior research addressed the problem by applying resampling methods such as the synthetic minority oversampling technique (SMOTE). However, resampling methods lead to other issues, e.g., increasing noisy data and training time during the process. To improve the bankruptcy prediction model, we propose cost-sensitive extreme gradient boosting (CS-XGB) to address the class imbalanced problem without requiring any resampling method. The proposed method’s effectiveness is evaluated on six real-world datasets, i.e., the LendingClub, and five Polish companies’ bankruptcy. This research compares the performance of CS-XGB with other ensemble methods, including SMOTE-XGB which applies SMOTE to the training set before the learning process. The experimental results show that i) based on LendingClub, the CS-XGB improves the performance of XGBoost and SMOTE-XGB by more than 50% and 33% on bankruptcy detection rate (BDR) and geometric mean (GM), respectively, and ii) the CS-XGB model outperforms random forest (RF), Bagging, AdaBoost, XGBoost, and SMOTE-XGB in terms of BDR, GM, and the area under a receiver operating characteristic curve (AUC) based on the five Polish datasets. Besides, the CS-XGB model achieves good overall prediction results.</span>

Predictive maintenance framework for assessing health state of centrifugal pumps

“…Similarly, feature selection and ensemble learning were additionally proposed in the economic sector. A light gradient boosting algorithm for risk analysis [50] and a cost sensitive-XGBoost approach for bankruptcy prediction [51] were developed with highly accurate results (around 95%) whilst [52] improved those outputs with an accuracy rate of 97%. On the other hand, principal component analysis (PCA) for feature selection [53] and Bayesian hyper-parameter optimization [54] were proposed in conjunction with XGBoost as optimal solutions for crowdfunding and credit worthiness prediction respectively.…”

Advancing Bankruptcy Forecasting With Hybrid Machine Learning Techniques: Insights From an Unbalanced Polish Dataset

“…However, models like LR and SVM can struggle with unbalanced datasets. Ensemble methods, including Random Forest, AdaBoost, Gradient Boosting, XGBoost, and CatBoost, have improved predictive performance, with Random Forest and XGBoost, in particular, being known for their accuracy and robustness in handling diverse datasets [13]- [16].…”

“…To address this, class-balancing methods like under-sampling, oversampling (including SMOTE and its variants), and combined techniques are frequently used. Some studies also explore unbalanced datasets without prior class balancing [7], [8], [13], and [24]. The ongoing development in bankruptcy prediction reflects the continuous pursuit of more accurate, efficient, and reliable methods to anticipate financial distress and mitigate its effects.…”

“…The dataset utilized in this study encompasses data on the bankruptcy of Polish companies. This dataset is well-known and extensively utilized in financial distress prediction research [9], [13], [24], and [28]. The data, which pertains to the probability of bankruptcy for Polish businesses, was sourced from the Emerging Markets Information Service (EMIS) database, providing details on global emerging markets.…”

“…A comparative analysis presented in Table VII between our proposed hybrid model and models from previous research, all applied to an unbalanced dataset, reveals that the This high performance was particularly notable when no feature selection was applied and parameters were optimized using Genetic Algorithms (GA). Notably, the CS-XGB model [13] achieved the second-highest AUC score of 0.955, despite not employing feature selection, parameter optimization, or data balancing techniques.…”