ICML
2001
IEEE
15 years 1 months ago
2001
IEEE
Accurate, well-calibrated estimates of class membership probabilities are needed in many supervised learning applications, in particular when a cost-sensitive decision must be mad...