Research Article
An Empirical Study on the Performance of Cost-Sensitive Boosting Algorithms with Different Levels of Class Imbalance
Algorithm 1
AdaBoost algorithm.
(i) Input: training set , where , BaseLearn | algorithm, Number of iterations . | (ii) Initialization: the weighted distribution of training samples: . | (iii) Iteration: For : | (1) Use the BaseLearn algorithm to train a component classifier on the training data set | sampling from distribution . | (2) Calculate the training error of the classifier : | (3) Set the weight for the classifier : | (4) Update the weighted distribution of training samples: | , | where is the normalization constant so that will be a distribution, that is, | . | (iv) Output: The final hypothesis: | |
|