Sparse semi-supervised heterogeneous interbattery bayesian analysis

作者：

Highlights：

• A novel heterogeneous multiview bayesian model for factor analysis.

• Include Sparse capabilities to automatically select the most relevant features.

• Work in a Semisupervised way to handle unlabelled data as well as missing values.

• It outperforms most of the state-of-the-art algorithms.

• Results show great versatility and an interpretability gain.

摘要

•A novel heterogeneous multiview bayesian model for factor analysis.•Include Sparse capabilities to automatically select the most relevant features.•Work in a Semisupervised way to handle unlabelled data as well as missing values.•It outperforms most of the state-of-the-art algorithms.•Results show great versatility and an interpretability gain.

论文关键词：Bayesian model,Canonical correlation analysis,Principal component analysis,Factor analysis,Feature selection,Semi-supervised,Multi-task

论文评审过程：Received 24 January 2020, Revised 15 March 2021, Accepted 28 June 2021, Available online 29 June 2021, Version of Record 15 July 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108141