Enterprise intelligent financial condition evaluation based on stacking ensemble model
The traditional methods for intelligent evaluation of the financial state of enterprises demonstrate low stability and poor cross-scenario adaptability of their results. This study proposes a new method for evaluating the financial state of enterprises based on the Stacked Generalization ensemble. It optimizes hyperparameters of the model and performs effective selection of financial features. It combines the Attention Mechanism with Weighted Fusion to achieve dynamic adaptation of heterogeneous financial data. Then it uses hierarchical fusion modeling based on Stacked Generalization to capture the internal relationships among financial indicators. In this way, it achieves efficient evaluation of enterprise financial conditions. In experiments, the model achieves an evaluation accuracy of 98.15% and a misclassification rate of 1.08%. Under cross-industry heterogeneous data scenarios, the scenario adaptation accuracy reaches 97.89%, and the evaluation consistency score reaches 98.23%. In enterprise investment and financing due diligence scenarios, the financial evaluation accuracy reaches 98.65%, and the financial risk tracing ability reaches 0.965. In the identification of four types of financial risks, the model achieves accuracies of 95.38%, 95.12%, 96.87%, and 96.63%. These results indicate that this method outperforms the comparative approaches in terms of evaluation precision, scenario adaptation, and risk recognition capability.
Authors
- Lan Shu (ORCID: https://orcid.org/0009-0001-9630-6842)
- Jiao Shu
- Qingqing Duan
Institutions
- Guangdong University Of Finances and Economics (CN)
- Hunan University of Technology (CN)
Publication Details
- Journal
- Discover Computing
- Published
- 2026-11-17
- DOI
- https://doi.org/10.1007/s10791-026-10638-4
- Primary Topic
- Financial Distress and Bankruptcy Prediction
- Type
- article
- Field-Weighted Citation Impact
- 0.00