Idea: use naïve Bayes on labeled examples and then apply EM. First, build naïve Bayes model on labeled data; Second, label unlabeled data based on class ...

好菇娘发布于2018/06/25

注脚

展开查看详情