Optimizing the Topology of Bayesian Network Classifiers by Applying Conditional Entropy to Mine Causal Relationships Between Attributes.
Limin WangGaojie WangZhiyi DuanHua LouMinghui SunPublished in: IEEE Access (2019)
Keyphrases
- causal relationships
- bayesian network classifiers
- conditional entropy
- bayesian networks
- information theory
- structure learning
- mutual information
- causal models
- discriminative learning
- naive bayes
- influence diagrams
- information theoretic
- data mining
- conditional independence
- class imbalance
- probabilistic model
- decision trees
- probability distribution
- naive bayes classifier
- data sets
- latent variables
- knn
- classification accuracy