Information resources estimation for accurate distribution-based concept drift detection
作者：
Highlights：
• Answers “How many labeled samples are required to detect concept drift accurately?”
• Analyses the information resources required for the detection of concept drift task.
• Proposed an analytical estimation framework for the concept drift detection task.
• Proposed the use of a lookup table for information resources estimation.
• Evaluated the proposed analytical framework on benchmarked data sets and methods.
摘要
•Answers “How many labeled samples are required to detect concept drift accurately?”•Analyses the information resources required for the detection of concept drift task.•Proposed an analytical estimation framework for the concept drift detection task.•Proposed the use of a lookup table for information resources estimation.•Evaluated the proposed analytical framework on benchmarked data sets and methods.
论文关键词：Concept drift,Information resources estimation,Probably Approximately Correct,Power Analysis,No Free Lunch theorem,Dynamic data stream
论文评审过程：Received 5 November 2021, Revised 26 January 2022, Accepted 18 February 2022, Available online 15 March 2022, Version of Record 15 March 2022.
论文官网地址：https://doi.org/10.1016/j.ipm.2022.102911
