Develops a MIP framework for joint optimization of scoring weights and risk category thresholds in interpretable healthcare tools, handling censored labels and asymmetric costs via constraints and a continuous relaxation.
Provably consistent partial-label learning.Advances in neural information processing systems, 33:10948–10960, 2020
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Joint Score-Threshold Optimization for Interpretable Risk Assessment
Develops a MIP framework for joint optimization of scoring weights and risk category thresholds in interpretable healthcare tools, handling censored labels and asymmetric costs via constraints and a continuous relaxation.