AutoWeka4MCPS-AVATAR: Accelerating automated machine learning pipeline composition and optimisation

作者：

Highlights：

• We propose the AVATAR to evaluate the pipeline validity using a surrogate model.

• Experiments show that invalid pipelines may result in overall bad performance.

• Experiments show the efficiency of multiple configuration initialisation of SMAC.

摘要

•We propose the AVATAR to evaluate the pipeline validity using a surrogate model.•Experiments show that invalid pipelines may result in overall bad performance.•Experiments show the efficiency of multiple configuration initialisation of SMAC.

论文关键词：Automated machine learning,Pipeline composition and optimisation,Machine learning pipeline evaluation,AutoML,Configuration space reduction

论文评审过程：Received 26 November 2020, Revised 8 April 2021, Accepted 19 July 2021, Available online 30 July 2021, Version of Record 4 August 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115643