An Accuracy-and-Diversity-based Ensemble Method for Concept Drift and Its application in Fraud Detection.
Shujie YinGuanjun LiuZhenchuan LiChungang YanChangjun JiangPublished in: ICDM (Workshops) (2020)
Keyphrases
- drifting concepts
- concept drift
- ensemble methods
- fraud detection
- prediction accuracy
- ensemble members
- ensemble classifier
- ensemble learning
- classifier ensemble
- data streams
- outlier detection
- non stationary
- benchmark datasets
- data mining techniques
- decision trees
- base classifiers
- classification algorithm
- random forest
- change detection
- cost sensitive
- random forests
- drift detection
- social network analysis
- class distribution
- data distribution
- machine learning methods
- data mining
- data sets
- multi class
- machine learning algorithms
- text mining
- classification accuracy
- machine learning