专利名称：DIVERSITY-AWARE WEIGHTED MAJORITY VOTE CLASSIFIER FOR IMBALANCED
DATASETS
台湾卤肉饭的做法
发明人：GOYAL, Anil,KHIARI, Jihed
av病毒申请号：EP2019/064900
申请日：20190606
性格分析测试公开号：WO2020/244771A1印度食物
公开日：
五月天亭亭
后妈也是妈
20201210
专利内容由知识产权出版社提供
专利附图：
摘要：The invention relates to an enmble learning bad method for a binary classification task on an imbalanced datat wherein the datat includes a minority class带树的成语
comprising positive samples and a majority class comprising negative samples. In order to reduce the fal positive rate of the classification the method compris generatively oversampling the datat by synthetically generating minority class examples, thereby generating datat , using the generated datat to generate subsamples and learning a ba classifier hi on each subsample , and learning a weighted majority vote classifier by combining the outputs of the ba classifiers , wherein each of the ba classifiers is assigned a weight in such a way that the diversity between the ba classifiers on the positive samples is minimized.
申请人：NEC LABORATORIES EUROPE GMBH
地址：Kurfürsten-Anlage 36 69115 Heidelberg DE喝椰汁的好处
国籍：DE
代理人：ULLRICH & NAUMANN
更多信息请下载全文后查看
