Financial distress prediction (FDP) models are effective tools to prevent stakeholders from suffering economic loss. In the process of FDP, the misclassification cost of typeⅠ error of the model is much higher than that of typeⅡerror. Some FPD models based on single classifiers take the asymmetric costs into consideration, but the study on cost-sensitive ensemble approach for FDP is rarely explored. This paper constructs cost-sensitive AdaBoost selective ensemble FDP model for minimizing

more »

... sification cost so that the loss of users of the model will suffer less. On the initial sample of 180 Chinese listed companies and 30 financial ratios, 8 times of holdout experiments are carried out for FDP respectively two years and three years in advance. The experimental results suggest that the proposed approach helps to reduce total misclassification costs compared with FDP model based on cost-sensitive C4.5 decision tree and that based on C4.5 decision tree.