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

A Framework Analysis for Alternating Components and Drift

  • Samuel Ackerman,
  • Guy Barash,
  • Eitan Farchi,
  • Orna Raz,
  • Onn Shehory

摘要

This chapter discusses a framework architecture that helps choose the best design of an ML system based on the data and models that compose the system. As data and business goals may change over time, the framework also supports offline analysis of data drift that causes performance degradation of the ML system.