Context-Aware Concept Distillation for Trustworthy Flood Prediction
用于可靠洪水预测的上下文感知概念蒸馏
机构 * School of Computer Science and Engineering, The Hebrew University of Jerusalem(耶路撒冷希伯来大学计算机科学与工程学院) ; Institute of Earth Sciences, The Hebrew University of Jerusalem(耶路撒冷希伯来大学地球科学研究所) ; Hebrew University Business School, The Hebrew University of Jerusalem(耶路撒冷希伯来大学商学院)
专题命中 安全评测 :trustworthy(title);safety(abstract);分类 cs.AI、cs.LG
AI总结 针对深度学习模型‘黑箱’问题阻碍洪水预测信任度的挑战,提出上下文感知概念蒸馏框架CACD,引入无监督管道和残差超网络,经全球多流域评估,该模型高保真且优于黑箱基线,平衡了AI准确性与决策透明度。
Comments to be published in IJCAI 2026 proceedings