Predikce bankrotu firem s využitím machine learning

The use of machine learning methods to create reliable models for predicting business failure is essential in today's challenging economic environment. Research focusing on data from the healthcare industry has shown that more complex machine learning methods are suitable for bankruptcy prediction. Complex algorithms provide better results in a dynamic, multi-factor financial modeling environment. Gradient boosting and random forest methods outperformed traditional Altman Z-score and Ohlson O-score models. The gradient boosting model features the highest values for key metrics such as accuracy, completeness and F1 score, effectively distinguishing between solvent and insolvent companies. The results suggest that a firm's financial health is affected by key factors such as liquidity, efficiency and capital structure.