2013b. Fast multi-instance multi-label learning

In multi-instance multi-label learning (MIML), one ob-ject is represented by multiple instances and simultane-ously associated with multiple labels. Existing MIML approaches have been found useful in many applica-tions; however, most of them can only handle moderate-sized data. To efficiently handle large data sets, we pro-pose the MIMLfast approach, which first constructs a low-dimensional subspace shared by all labels, and then trains label specific linear models to optimize approxi-mated ranking loss via stochastic gradient descent. Al-though the MIML problem is complicated, MIMLfast is able to achieve excellent performance by exploit-ing label relations with shared space and discovering sub-concepts for complicated labels. Experiments show that the performance of MIMLfast is highly competitive to state-of-the-art techniques, whereas its time cost is much less; particularly, on a data set with 30K bags and 270K instances, where none of existing approaches can return results in 24 hours, MIMLfast takes only 12 min-utes. Moreover, our approach is able to identify the most representative instance for each label, and thus provid-ing a chance to understand the relation between input patterns and output semantics