Clustered Calibration: Representation-Aware Probability Calibration via Learned Subpopulations

ArXi:2510.19328v2 Announce Type: replace Ensuring that predicted probabilities align with observed frequencies is critical in high-stakes domains such as clinical decision, autonomous driving and financial risk assessment. Existing calibration methods typically apply a single global transformation or rely on post-hoc binning over predicted confidences, limiting their ability to exploit heterogeneous reliability across sub-populations.