Circular Bayesian classifiers using wrapped Cauchy distributions
作者：
Highlights：
• An adaptation of Bayesian network classifiers to circular domain is proposed.
• The classifiers are suitable for wrapped Cauchy distributions.
• A conditional circular mutual information for wrapped Cauchy is presented
• Wrapped Cauchy classifiers outperform linear Bayesian network classifiers.
• Neuron layer identification is appropriate for wrapped Cauchy classifiers.
摘要
•An adaptation of Bayesian network classifiers to circular domain is proposed.•The classifiers are suitable for wrapped Cauchy distributions.•A conditional circular mutual information for wrapped Cauchy is presented•Wrapped Cauchy classifiers outperform linear Bayesian network classifiers.•Neuron layer identification is appropriate for wrapped Cauchy classifiers.
论文关键词：Data mining,Classification,Circular statistics,Wrapped Cauchy distribution,Bayesian networks,Cortical layer
论文评审过程：Received 23 October 2017, Revised 24 November 2018, Accepted 25 May 2019, Available online 28 May 2019, Version of Record 25 July 2019.
论文官网地址：https://doi.org/10.1016/j.datak.2019.05.005
