错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Large-Scale Data Challenges: Instability in Statistical Learning

  • Bo-Yu Chen,
  • Hao Zhang

摘要

Numerous approximation methods have been developed to approximate both the kernel matrix and its inverse. We investigate one such influential approximation that has recently gained popularity. However, our results indicate that this approximation fails to address the ill-conditioning of the kernel matrix, potentially leading to significantly large biases and highly unstable prediction results.