发表机构
Université Paris-Saclay; CNRS; Laboratoire Interdisciplinaire des Sciences du Numérique（巴黎-萨克雷大学; 法国国家科学研究中心; 跨学科数字科学实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究揭示自适应分类器内部更新与外部漂移检测的冲突，提出盲点现象，并量化其临界条件，证明经典基准因幅度不足而无法触及。
AI 中文摘要
从自适应分类器的错误流中监测概念漂移，会与模型自身的更新循环产生操作冲突。当内部自适应速度超过证据积累速度时，准确率在累积检测器（CUSUM、Page-Hinkley）达到阈值之前就已恢复。对自适应随机森林（ARF）的检测表明，幸存的树仅通过增量叶更新就吸收了漂移后错误瞬态的98.6%。第一次背景树交换仅占被消除错误量的0.71%，却使外部检测率下降了31个百分点。我们推导了累积证据无法跨越阈值的有限时域边界，并测量了一个临界幅度下限（$\Delta e_c = 0.120$），低于该下限时误报预算会阻止检测。这种失效表现为静态流上被遗漏的漂移，以及噪声基线上由内部树交换引发的误报泛滥。我们在合成漂移、ARMA-GARCH序列（ProteuS）和表格基准（BAF、INSECTS）上进行了验证；在标准阈值下的合成扫描中，盲点出现在$\Delta e \approx 0.25$，这解释了为什么像SEA（$\Delta e \le 0.21$）这样的经典基准未能达到该盲点。
英文摘要
Monitoring concept drift from an adaptive classifier's error stream creates an operational conflict with the model's own update loop. When internal adaptation outpaces evidence accumulation, accuracy recovers before cumulative detectors (CUSUM, Page-Hinkley) can reach threshold. Instrumenting an Adaptive Random Forest (ARF) shows that surviving trees absorb 98.6% of the post-drift error transient through incremental leaf updates alone. The first background tree swap accounts for just 0.71% of this erased error volume, but drops external detection rates by 31 percentage points. We derive the finite-horizon boundary where cumulative evidence fails to cross threshold and measure a critical magnitude floor ($Δe_c = 0.120$) below which false-alarm budgets preclude detection. This failure manifests as missed shifts on stationary streams and false-alarm flooding triggered by internal tree swaps on noisy baselines. We validate on synthetic shifts, ARMA-GARCH series (ProteuS), and tabular benchmarks (BAF, INSECTS); on the synthetic sweep at a standard threshold, the blind spot appears at $Δe \approx 0.25$, showing why classical benchmarks like SEA ($Δe \le 0.21$) failed to reach it.
CommentsAccepted at IEEE ICDM 2026 Workshops (OWAD 2026). 10 pages, 3 figures, 2 tables. Code: https://github.com/TheBlindSpotParadox/TheBlindSpotParadox-Experiments
