主题:QUIRE,基于active-learning的最小-最大原则,实现信息量大而具有代表性的数据筛选。
Informativeness: measures the ability of an instance in reducing the uncertainty of a statistical model.忽视了数据间的潜在结构分布。
Representativeness: measures if an instance well represents the overall input patterns of unlabeled data.需要用到大量数据,依赖于聚类效果