The Bayesian logistic regression in pattern recognition problems under concept drift.
Pavel A. TurkovOlga KrasotkinaVadim MottlPublished in: ICPR (2012)
Keyphrases
- pattern recognition problems
- concept drift
- bayesian logistic regression
- high order
- predictive power
- data streams
- pattern recognition
- streaming data
- drifting concepts
- change detection
- non stationary
- drift detection
- classification algorithm
- data distribution
- combination of multiple classifiers
- higher order
- batch learning
- ensemble learning
- data stream classification
- predictive accuracy
- logistic regression
- prediction accuracy
- neural network
- pairwise
- ensemble classifier
- image processing
- decision trees
- support vector
- active learning
- machine learning