A hierarchical multiple kernel support vector machine for customer churn prediction using longitudinal behavioral data.
Zhen-Yu ChenZhi-Ping FanMinghe SunPublished in: Eur. J. Oper. Res. (2012)
Keyphrases
- multiple kernel
- support vector machine
- customer churn
- behavioral data
- prediction model
- churn prediction
- kernel methods
- prediction accuracy
- machine learning
- multiple kernel learning
- multi class
- cognitive model
- svm classifier
- feature selection
- kernel function
- feature vectors
- support vector machine svm
- training set
- support vector
- linear combination
- behavior patterns
- training data
- raw data
- regression model
- kernel learning
- cost sensitive
- semi supervised
- random forest
- knn
- feature space