ICML
2010
IEEE
2010
IEEE
Unsupervised Risk Stratification in Clinical Datasets: Identifying Patients at Risk of Rare Outcomes
14 years 3 months ago
Most existing algorithms for clinical risk stratification rely on labeled training data. Collecting this data is challenging for clinical conditions where only a small percentage ...
ACSC
2005
IEEE
14 years 8 months ago
2005
IEEE
Most current network intrusion detection systems employ signature-based methods or data mining-based methods which rely on labelled training data. This training data is typically ...