A Nonparametric Approach to Uncovering Connected Anomalies by Tree Shaped Priors.
Nannan WuFeng ChenJianxin LiJinpeng HuaiBaojian ZhouBo LiNaren RamakrishnanPublished in: IEEE Trans. Knowl. Data Eng. (2019)
Keyphrases
- tree shaped
- parametric models
- bayesian approaches
- anomaly detection
- data driven
- dirichlet process
- bayesian modeling
- information systems
- similarity measure
- bayesian framework
- anomalous behavior
- density estimation
- change point detection
- kernel density estimation
- database
- prior information
- connected components
- learning algorithm