Multiple Instance Learning with the Optimal Sub-Pattern Assignment Metric
This work addresses the challenge of processing multiple instance data for researchers and practitioners in machine learning, though it appears incremental as it adapts an existing metric to this domain.
The paper tackled the problem of handling unordered sets or multi-sets in multiple instance learning tasks like clustering, classification, and novelty detection by introducing the Optimal Sub-Pattern Assignment metric, with numerical experiments on simulated and real data demonstrating its versatility.
Multiple instance data are sets or multi-sets of unordered elements. Using metrics or distances for sets, we propose an approach to several multiple instance learning tasks, such as clustering (unsupervised learning), classification (supervised learning), and novelty detection (semi-supervised learning). In particular, we introduce the Optimal Sub-Pattern Assignment metric to multiple instance learning so as to provide versatile design choices. Numerical experiments on both simulated and real data are presented to illustrate the versatility of the proposed solution.