Early identification of ICU patients at risk of complications: Regularization based on robustness and stability of explanations.
Tiago AmadorSaulo SaturninoAdriano VelosoNivio ZivianiPublished in: Artif. Intell. Medicine (2022)
Keyphrases
- high risk
- intensive care unit
- intensive care units
- risk factors
- empirical risk minimization
- clinical data
- prostate cancer
- medical staff
- empirical risk
- cardiac surgery
- intensive care
- blood pressure
- stability analysis
- acute coronary syndrome
- numerical stability
- monitoring system
- decision making
- risk assessment
- causal models
- vital signs
- risk management
- prior information
- expert systems
- generating explanations
- decision trees