Its weighted training error and scaling factor cm is computed (step 3b).The weights are increased for training samples, which have been misclassified (step 3c).
英
美
- 计算错误率和换算系数cm(step 3b).;被错分的样本的权重会增加。