ABSTRACT
In this paper, a new framework for unsupervised partitioning problems are proposed in order to stabilize the performance of the Large-Margin Metric Learning method developed by Lajugie [lajugie et al. 2014] when the fraction of the given labels is low. The Large-Margin method could be used to deal with a wide range of partitioning problems such as clustering, image segmentation, video segmentation and change-detection problems but does not show robustness when applied to large set of data with a small number of labels. Hence, we combine the algorithm with the relevant component analysis algorithm [Bar-Hillel et al.2006] when the fraction of the labels is low and adhere to the original Large-Margin Method when the fraction is high. We also provide experiments to show that by implementing the new frame work, we can achieve more stable partitioning performance in synthetic examples.
- R. Lajugie, F. Bach, and S. Arlot, "Large margin metric learning for constrained partitioning problems," in Proc. International Conferenceon Machine Learning, 2014. Google ScholarDigital Library
- A. Bar-Hillel, T. Hertz, N. Shental, and D. Weinshall, "Learning a mahalanobismetric from equivalence constraints." Journal of Machine Learning Research, vol. 6, pp. 937--965,2005. {Online}. Available:http://dblp.unitrier.de/db/journals/jmlr/jmlr6.html-Bar-HillelHSW05 Google ScholarDigital Library
- E. P. Xing, M. I. Jordan, S. J. Russell, and A. Y.Ng, "Distance metric learning with application to clustering with side-information," pp. 521--528,2003. {Online}. Available:http://papers.nips.cc/paper/2164-distance-metric-learning-withapplication-to-clustering-with-side-information.pdf Google ScholarDigital Library
- E. R. Killick, P. Fearnhead, "Optimal detection of change points with a linear computational cost" Journal of the American Statistical Association, vol. 107, no. 500, pp. 1590--1598, 2012. {Online}.Available:Google ScholarCross Ref
- E. Forgy, "Cluster analysis of multivariate data: efficiency versus interpretability of classifications," Biometrics, vol. 21, pp. 768--780,1965.Google Scholar
- F. R. Bach and M. I. Jordan, "Learning spectral clustering,"no. UCB/CSD-03-1249, Jun 2003. {Online}. Available: http://www.eecs.berkeley.edu/Pubs/TechRpts/2003/5549.htmlGoogle Scholar
- T. Finley and T. Joachims, "Supervised clustering with support vector machines," pp. 217--224, 2005. {Online}. Available:http://portal.acm.org/citation.cfm?id=1102351.110 2379-----, "Supervised k-means clustering," 2008. Google ScholarDigital Library
- K. Q. Weinberger, J. Blitzer, and L. K. Saul, "Distance metric learning for large margin nearest neighbor classification," pp. 1473--1480, 2006. {Online}. Available: http://papers.nips.cc/paper/2795-distancemetric-learning-for-large-margin-nearest-neighbor-classification.pdf Google ScholarDigital Library
- M. Szummer, "Learning CRFs using graph cuts," in ECCV, October 2008. {Online}. Available: https://www.microsoft.com/enus/research/publication/learning-CRFs-using-graph-cuts/ Google ScholarDigital Library
- Y. Wang and S. Chen, "Soft large margin clustering." Inf. Sci., vol. 232, pp. 116--129, 2013. {Online}. Available: http://dblp.unitrier.de/db/journals/isci/isci232.html-WangC13 Google ScholarDigital Library
- D. M. Johnson, C. Xiong, and J. J. Corso, "Semi-supervised nonlinear distance metric learning via forests of max-margin cluster hierarchies." IEEE Trans. Knowl. Data Eng., vol. 28,no. 4, pp. 1035--1046, 2016. {Online}. Available: http://dblp.unitrier.de/db/journals/tkde/tkde28.html-JohnsonXC16 Google ScholarDigital Library
- F. Wang, R. Li, Z. Lei, X. S. Ni, X. Huo, and M. Chen, "Kernel fusion--refinement for semi-supervised nonlinear dimension reduction," Pattern Recognition Letters, vol. 63, pp. 16--22,2015. Google ScholarDigital Library
- B. Shi, Z.-F. Pang, and J. Xu, "Image segmentation based on the hybrid total variation model and the k-means clustering strategy,"arXiv preprint arXiv: 1605.09116, 2016.Google Scholar
- B. Kulis and M. I. Jordan, "Revisiting k-means: New algorithms via bayesian nonparametrics," arXiv preprint arXiv:1111.0352, 2011.Google Scholar
- P. Jain, B. Kulis, J. Davis, and I. Dhillon. 2012. Metric and kernel learning using a linear transformation. Journal of Machine Learning Research 13 (3) 519--547, 2012. Google ScholarDigital Library
- A. Joulin, F. Bach, and J. Ponce. Discriminative clustering for image co-segmentation. In Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on. IEEE, 1943--1950,2010.Google Scholar
- B. T. C. G. D. Roller, "Max-margin Markov networks," Advances in neural information processing systems, vol. 16, p. 25, 2004.Google Scholar
- I. Tsochantaridis, T. Hofmann, T. Joachims, and Y. Altun, "Support vector machine learning for interdependent and structured output spaces" p. 104, 2004. Google ScholarDigital Library
Recommendations
Maximum margin partial label learning
Partial label learning aims to learn from training examples each associated with a set of candidate labels, among which only one label is valid for the training example. The basic strategy to learn from partial label examples is disambiguation, i.e. by ...
Large margin metric learning for multi-label prediction
AAAI'15: Proceedings of the Twenty-Ninth AAAI Conference on Artificial IntelligenceCanonical correlation analysis (CCA) and maximum margin output coding (MMOC) methods have shown promising results for multi-label prediction, where each instance is associated with multiple labels. However, these methods require an expensive decoding ...
A large margin algorithm for automated segmentation of white matter hyperintensity
A novel large margin method for white matter hyperintensity segmentation is proposed.A supervised large margin algorithm is proposed to learn a global classifier.A semi-supervised large margin classifier is learned for refinement on test subject.The ...
Comments