Tilted Mondrian：改进保形预测中的细粒度公平性
Tilted Mondrian: Triage and Mitigation of Fine-grained Disparities in Conformal Prediction
如果你用保形预测做过分组校准，这篇指出了 Mondrian 方法在细分层里覆盖率的坑，Tilted Mondrian 不用重训模型就能缓解，值得看方法部分。
论文针对 Mondrian 校准在保形预测中的公平性问题：组级别的覆盖率保证过于粗糙，更细粒度的分层内仍可能存在显著差异。作者提出倾向标准化方法来评估人群构成如何导致粗粒度与细粒度比较之间的不一致。缓解方案 Tilted Mondrian 无需额外模型训练，在保留组级覆盖率保证的同时缩小细粒度分层内的覆盖率和预测集大小差异。论文还推导了一个与方法无关的界，量化细粒度覆盖率差异最多能减少多少。
Tilted Mondrian: Triage and Mitigation of Fine-grained Disparities in Conformal Prediction
Fairness in conformal prediction is often pursued through Mondrian calibration, which provides finite-sample coverage guarantees separately across groups. These guarantees, however, are coarse: substantial disparities can remain within finer strata. This coarse-to-fine fairness issue raises two questions: how severe can such disparities be, and how can they be mitigated? Aggregate disparities can depend on how groups are distributed across strata. For severity assessment, we introduce propensity standardization to examine the effect of population composition, and show how group composition can induce disagreement between aggregate and finer-stratum comparisons. For mitigation, we develop Tilted Mondrian to improve fairness compared to group-Mondrian at a finer level without additional model training. Tilted Mondrian retains groupwise coverage guarantees and mitigates coverage and prediction set size disparities within finer strata. Finally, we derive a method-agnostic bound that quantifies how much reduction in finer-stratum coverage disparity can be achieved.
