Feature Ranking from Random Forest Through Complex Network's Centrality Measures - A Robust Ranking Method Without Using Out-of-Bag Examples.
Adriano Henrique CantãoAlessandra Alaniz MacedoLiang ZhaoJosé Augusto BaranauskasPublished in: ADBIS (2022)
Keyphrases
- random forest
- complex networks
- feature ranking
- centrality measures
- social network analysis
- network analysis
- decision trees
- fold cross validation
- clustering coefficient
- feature set
- social networks
- feature selection
- multi label
- ensemble learning
- link prediction
- network structure
- graph theory
- community structure
- ensemble methods
- training examples
- small world
- base classifiers
- network topology
- text mining
- feature vectors
- feature subset
- neural network
- genetic algorithm
- machine learning