A mixed integer programming algorithm for minimizing the training sample misclassification cost in two-group classification.
António Pedro Duarte SilvaAntonie StamPublished in: Ann. Oper. Res. (1997)
Keyphrases
- training samples
- mixed integer programming
- learning algorithm
- similarity measure
- classification algorithm
- classification accuracy
- np hard
- cost sensitive
- optimal solution
- training set
- dynamic programming
- support vector machine
- decision trees
- support vector machine svm
- multi class
- supervised learning
- expectation maximization
- model selection
- upper bound
- support vector
- objective function
- mixed integer
- misclassification costs
- training data