Bag Reference Vector for Multi-instance Learning
This addresses the challenge of ambiguous instance labels in positive bags for multi-instance learning applications, representing an incremental improvement over existing methods.
The paper tackles the problem of multi-instance learning by proposing a method to describe each bag using feature vectors based on similarities with other bags, rather than focusing on instances alone, and reports that it outperforms previous state-of-the-art methods by a large margin in benchmarks and text categorization tasks.
Multi-instance learning (MIL) has a wide range of applications due to its distinctive characteristics. Although many state-of-the-art algorithms have achieved decent performances, a plurality of existing methods solve the problem only in instance level rather than excavating relations among bags. In this paper, we propose an efficient algorithm to describe each bag by a corresponding feature vector via comparing it with other bags. In other words, the crucial information of a bag is extracted from the similarity between that bag and other reference bags. In addition, we apply extensions of Hausdorff distance to representing the similarity, to a certain extent, overcoming the key challenge of MIL problem, the ambiguity of instances' labels in positive bags. Experimental results on benchmarks and text categorization tasks show that the proposed method outperforms the previous state-of-the-art by a large margin.