Clairvoyant: AdaBoost with Cost-Enabled Cost-Sensitive Classifier for Customer Churn Prediction

Table 3

Reweighting step of discrete AdaBoost and proposed AdaBoostWithCost.

Discrete AdaBoost	AdaBoostWithCost
Update weight step:	Update weight step:
(where )
(where )

is the new probability assigned to the data point at iteration; all other parameters constituting the right side of the equation are described as follows:  = hypopaper’s weight for gross misclassification error at iteration,  = false-negative error at iteration,  = misclassification cost for false-negative error specified in the input cost matrix,.