A framework for expert-driven subpopulation discovery and evaluation using subspace clustering for epidemiological data
作者：
Highlights：
• Framework for knowledge discovery, model inspection and model validation.
• Model learning is automated and self-tuned.
• Semi-supervised selection of variables with minimal information from the expert.
• Mechanisms to inspect and validate models on independent cohort data.
• Investigation of the framework’s potential on outcomes hepatic steatosis and goiter.
摘要
•Framework for knowledge discovery, model inspection and model validation.•Model learning is automated and self-tuned.•Semi-supervised selection of variables with minimal information from the expert.•Mechanisms to inspect and validate models on independent cohort data.•Investigation of the framework’s potential on outcomes hepatic steatosis and goiter.
论文关键词：Subpopulation discovery framework,Constraint-based subspace clustering,Cohort study data,Hepatic steatosis,Goiter
论文评审过程：Received 19 September 2017, Revised 11 June 2018, Accepted 2 July 2018, Available online 2 July 2018, Version of Record 9 July 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.07.003
