arXiv 机器学习· Yukai Song (Department of Electrical and Computer Engineering, University of Pittsburgh), Yangfan Deng (Department of Electrical and Computer Engineering, University of Maryland, College Park), Jijun Yin (Department of Electrical and Computer Engineering, University of Pittsburgh), Zhi-Hong Mao (Department of Electrical and Computer Engineering, University of Pittsburgh), Jingtong Hu (Department of Electrical and Computer Engineering, University of Pittsburgh)·· 4 小时前AI 评分31
COVER：在选择性睡眠分期中提高模型覆盖率
COVER: Learning to Accept More in Selective Sleep Staging
AI 导读
研究人员提出选择性睡眠分期方法 COVER，通过辅助误差学习与固定尺度评分细化，在受控风险约束下最大化初级分类器的预测覆盖率。在 Sleep-EDF-20 的 5% 风险目标测试中，COVER 在 4.5% 风险下取得 50.4% 的最高平均覆盖率，优于基准 SELE（49.3%）和 DuoF（43.6%）。该方法还在同等接受量下将错误数从传统 MSP 的 1,467 降至 867。
来源：arXiv 机器学习 · arxiv.org