A new hybrid classification algorithm for customer churn prediction based on logistic regression and decision trees.
Arno De CaignyKristof CoussementKoen W. De BockPublished in: Eur. J. Oper. Res. (2018)
Keyphrases
- classification algorithm
- logistic regression
- naive bayes
- decision trees
- customer churn
- churn prediction
- prediction model
- rule induction
- random forest
- training set
- cost sensitive
- classification accuracy
- random forests
- credit scoring
- classification rules
- prediction accuracy
- training data
- classification trees
- breast cancer
- knn
- ensemble methods
- machine learning algorithms
- text categorization
- text classification
- customer relationship management
- logistic regression models
- probability estimates
- learning algorithm
- base classifiers
- decision rules
- regression model
- support vector machine
- machine learning
- data mining methods
- k nearest neighbor
- survival analysis
- support vector
- bayesian classifiers
- feature selection
- knowledge discovery
- feature extraction
- prognostic models