ABSTRACT
In traditional multi-instance learning (MIL), instances are typically represented by using a single feature view. As MIL becoming popular in domain specific learning tasks, aggregating multiple feature views to represent multi-instance bags has recently shown promising results, mainly because multiple views provide extra information for MIL tasks. Nevertheless, multiple views also increase the risk of involving redundant views and irrelevant features for learning. In this paper, we formulate a new cross-view feature selection problem that aims to identify the most representative features across all feature views for MIL. To achieve the goal, we design a new optimization problem by integrating both multi-view representation and multi-instance bag constraints. The solution to the objective function will ensure that the identified top-m features are the most informative ones across all feature views. Experiments on two real-world applications demonstrate the performance of the cross-view feature selection for content-based image retrieval and social media content recommendation.
- S. Andrews, I. Tsochantaridis, and T. Hofmann. Support vector machines for multiple-instance learning. In NIPS, pages 561--568, 2003.Google ScholarDigital Library
- S.-F. Chang. Segmentation using superpixels: A bipartite graph partitioning approach. In CVPR, pages 789--796, 2012. Google ScholarDigital Library
- C. Christoudias, R. Urtasun, and T. Darrell. Unsupervised feature selection via distributed coding for multi-view object recognition. In CVPR, pages 1--8, 2008.Google ScholarCross Ref
- Z. Fang and Z. M. Zhang. Discriminative feature selection for multi-view cross-domain learning. In CIKM, pages 1321--1330, 2013. Google ScholarDigital Library
- R. Gan and J. Yin. Feature selection in multi-instance learning. Neural Comput. Appl., 23(3-4):907--912, 2013.Google ScholarCross Ref
- M. Grbovic, C. Dance, and S. Vucetic. Sparse principal component analysis with constraints. In AAAI, pages 935--941, 2012.Google Scholar
- Z. Hong, X. Mei, D. Prokhorov, and D. Tao. Tracking via robust multi-task multi-view joint sparse representation. In ICCV, pages 649--656, 2013. Google ScholarDigital Library
- X. Kong and P. S. Yu. Semi-supervised feature selection for graph classification. In KDD, pages 793--802, 2010. Google ScholarDigital Library
- H. Lakkaraju, J. McAuley, and J. Leskovec. What's in a name? understanding the interplay between titles, content, and communities in social media. In ICWSM, pages 311--320, 2013.Google Scholar
- J. Li and J. Z. Wang. Real-time computerized annotation of pictures. IEEE Trans. Pattern Anal. Mach. Intell., 30:985--1002, 2008. Google ScholarDigital Library
- M. Mayo and E. Frank. Experiments with multi-view multi-instance learning for supervised image classification. In IVCNZ, pages 363--369, 2011.Google Scholar
- C. Nguyen, X. Wang, J. Liu, and Z.-H. Zhou. Labeling complicated objects: Multi-view multi-instance multi-label learning. In AAAI, 2014.Google Scholar
- S. Pan, X. Zhu, C. Zhang, and P. S. Yu. Graph stream classification using labeled and unlabeled graphs. In ICDE, pages 398--409, 2013. Google ScholarDigital Library
- W. Ping, Y. Xu, K. Ren, H. Chi, and S. Furao. Non-i.i.d. multi-instance dimensionality reduction by learning a maximum bag margin subspace. In AAAI, pages 551--556, 2010.Google Scholar
- S. Ray and M. Craven. Supervised versus multiple instance learning: an empirical comparison. In ICML, pages 697--704, 2005. Google ScholarDigital Library
- A. Sun. Short text classification using very few words. In SIGIR, pages 1145--1146, 2012. Google ScholarDigital Library
- J. Tang, X. Hu, H. Gao, and H. Liu. Unsupervised feature selection for multi-view data in social media. In SDM, pages 270--278, 2013.Google ScholarCross Ref
- K. E. A. Van de Sande, T. Gevers, and C. G. M. Snoek. Evaluating color descriptors for object and scene recognition. IEEE Trans. Pattern Anal. Mach. Intell., 32(9):1582--1596, 2010. Google ScholarDigital Library
- J. Wu, Z. Hong, S. Pan, X. Zhu, C. Zhang, and Z. Cai. Multi-graph learning with positive and unlabeled bags. In SDM, pages 217--225, 2014.Google ScholarCross Ref
- J. Wu, X. Zhu, C. Zhang, and Z. Cai. Multi-instance multi-graph dual embedding learning. In ICDM, pages 827--836, 2013.Google ScholarCross Ref
- J. Wu, X. Zhu, C. Zhang, and P. Yu. Bag constrained structure pattern mining for multi-graph classification. IEEE Trans. on Knowl. and Data Eng., PP(99):1--1, 2014.Google ScholarCross Ref
- T. Xia, D. Tao, T. Mei, and Y. Zhang. Multiview spectral embedding. Trans. Sys. Man Cyber. Part B, 40:1438--1446, 2010. Google ScholarDigital Library
- B. Xie, Y. Mu, D. Tao, and K. Huang. m-sne: Multiview stochastic neighbor embedding. Trans. Sys. Man Cyber. Part B, 41(4):1088--1096, 2011. Google ScholarDigital Library
- J. Yu, D. Liu, D. Tao, and H. S. Seah. On combining multiple features for cartoon character retrieval and clip synthesis. IEEE Trans. Sys. Man Cyber. Part B, 42(5):1413--1427, 2012.Google ScholarDigital Library
- H. Yuan, M. Fang, and X. Zhu. Hierarchical sampling for multi-instance ensemble learning. IEEE Trans. on Knowl. and Data Eng., 25(12):2900--2905, 2013. Google ScholarDigital Library
- A. Zafra, M. Pechenizkiy, and S. Ventura. Relieff-mi: An extension of relieff to multiple instance learning. Neurocomput., 75(1):210--218, 2012. Google ScholarDigital Library
- D. Zhang, J. He, and R. Lawrence. Mi2ls: Multi-instance learning from multiple information sources. In KDD, pages 149--157, 2013. Google ScholarDigital Library
- M.-L. Zhang and Z.-H. Zhou. Improve multi-instance neural networks through feature selection. Neural Process. Lett., 19(1):1--10, 2004. Google ScholarDigital Library
- Z.-H. Zhou, Y.-Y. Sun, and Y.-F. Li. Multi-instance learning by treating instances as non-i.i.d. samples. In ICML, pages 1249--1256, 2009. Google ScholarDigital Library
- Z.-H. Zhou, M.-L. Zhang, S.-J. Huang, and Y.-F. Li. Multi-instance multi-label learning. Artif. Intell., 176:2291--2320, 2012. Google ScholarDigital Library
Index Terms
- Exploring Features for Complicated Objects: Cross-View Feature Selection for Multi-Instance Learning
Recommendations
Labeling complicated objects: multi-view multi-instance multi-label learning
AAAI'14: Proceedings of the Twenty-Eighth AAAI Conference on Artificial IntelligenceMulti-Instance Multi-Label (MIML) is a learning framework where an example is associated with multiple labels and represented by a set of feature vectors (multiple instances). In the formalization of MIML learning, instances come from a single source (...
Cost‐effective multi‐instance multilabel active learning
AbstractMulti‐instance multi‐label (MIML) Active Learning (M2AL) aims to improve the learner while reducing the cost as much as possible by querying informative labels of complex bags composed of diverse instances. Existing M2AL solutions suffer high ...
Diversified dictionaries for multi-instance learning
Multiple-instance learning (MIL) has been a popular topic in the study of pattern recognition for years due to its usefulness for such tasks as drug activity prediction and image/text classification. In a typical MIL setting, a bag contains a bag-level ...
Comments