提出自适应温度校准法，让分类概率更可靠且有统计保证。
Adaptive Cumulative Mass Calibration with Conformal Prediction
- 基于分位数校准思想，用置信区间约束概率输出。
- 在多类别图像任务中，新指标和传统指标均优于现有方法。
- 适合对可靠性要求高的场景，如医疗诊断与自动驾驶。
分类器提供可靠的概率估计在高风险应用中至关重要。然而实际中预测概率常存在校准偏差，现有后处理校准方法通常无法保证特定校准性态的达成。本文通过累积质量校准（cumulative mass calibration）及其误差度量，引入一种集合视角的校准方法。提出基于符合预测的新校准流程，可保证边际覆盖性。设计了自适应温度缩放算法，为每个输入动态调整温度以满足覆盖约束。该方法可高效实现。在图像分类任务中，尤其在多类别设置下，本方法在新提出的校准误差度量（CMCE 和 α-CMCE）以及标准指标（如 ECE、cw-ECE、MCE）上均优于现有基线。
原文摘要 · Abstract (English)
Reliable probability estimates by classifiers are essential in high-risk applications. In practice, however, predicted probabilities are often miscalibrated, and many existing post-hoc calibration methods typically lack guarantees that a specific notion of calibration is achieved after the correction procedure is applied. We introduce a *set-based* perspective on calibration through the notion of *cumulative mass calibration* and the corresponding error measures. We propose a new calibration procedure based on conformal prediction that forms cumulative probabilities with guaranteed marginal coverage. We introduce an __adaptive temperature scaling algorithm__, with the temperature tuned for each input to satisfy the conformal coverage constraint. As we show, this procedure can be efficiently implemented. Across image classification tasks, particularly in settings with many classes, our method improves newly introduced calibration error measures (__CMCE__ and $α$-CMCE) *and* standard metrics (such as ECE, cw-ECE, MCE) over the existing baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。