认识到在处理相互冲突的多视图数据时显式提供决策结果和相关可靠性的重要性,指出一个新的可靠冲突多视图学习 (RCML) 问题。 提出了一种针对 RCML 问题的证据冲突多视图学习 (ECML) 方法。 在多视图融合阶段,提出一种冲突意见聚合策略,并从理论上证明了该策略可以精确地模拟多视图公共和视图特定可靠性的关系。 总体...
Multi-view learning aims to combine multiple features to achieve more comprehensive descriptions of data. Most previous works assume that multiple views are strictly aligned. However, real-world multi-view data may contain low-quality conflictive instances, which show conflictive information in different...
In the literature on multi-view learning techniques, "view disagreement" is listed as among the major challenges for multi-view representation learning. View disagreement is distinct from nonshared features in alternative views, because it gives rise to a type of uncertainty that humans find ...