ABSTRACT
Most traditional supervised learning methods are developed to learn a model from labeled examples and use this model to classify the unlabeled ones into the same label space predefined by the models. However, in many real world applications, the label spaces for both the labeled/training and unlabeled/testing examples can be different. To solve this problem, this paper proposes a novel notion of Serendipitous Learning (SL), which is defined to address the learning scenarios in which the label space can be enlarged during the testing phase. In particular, a large margin approach is proposed to solve SL. The basic idea is to leverage the knowledge in the labeled examples to help identify novel/unknown classes, and the large margin formulation is proposed to incorporate both the classification loss on the examples within the known categories, as well as the clustering loss on the examples in unknown categories. An efficient optimization algorithm based on CCCP and the bundle method is proposed to solve the optimization problem of the large margin formulation of SL. Moreover, an efficient online learning method is proposed to address the issue of large scale data in online learning scenario, which has been shown to have a guaranteed learning regret. An extensive set of experimental results on two synthetic datasets and two datasets from real world applications demonstrate the advantages of the proposed method over several other baseline algorithms. One limitation of the proposed method is that the number of unknown classes is given in advance. It may be possible to remove this constraint if we model it by using a non-parametric way. We also plan to do experiments on more real world applications in the future.
- J. Allan. Topic detection and tracking: event-based information organization. Kluwer Academic Publishers Norwel, 2002. Google ScholarDigital Library
- K. P. Bennett and A. Demiriz. Semi-supervised support vector machines. In NIPS, pages 368--374, 1998. Google ScholarDigital Library
- J. Betteridge, A. Carlson, S. Hong, and E. Hruschka Jr. Toward Never Ending Language Learning. In AAAI, 2009.Google Scholar
- A. Carlson, J. Betteridge, B. Kisiel, B. Settles, E. R. H. Jr., and T. M. Mitchell. Toward an architecture for never-ending language learning. In AAAI, 2010.Google ScholarDigital Library
- R. Duda, P. Hart, and D. Stork. Pattern classification. Wiley-Interscience, 2001. Google ScholarDigital Library
- E. Hazan, A. Kalai, S. Kale, and A. Agarwal. Logarithmic regret algorithms for online convex optimization. In COLT, pages 499--513, 2006. Google ScholarDigital Library
- T. Joachims. Transductive inference for text classification using support vector machines. In ICML, pages 200--209, 1999. Google ScholarDigital Library
- D. D. Lewis, Y. Yang, T. G. Rose, and F. Li. Rcv1: A new benchmark collection for text categorization research. Journal of Machine Learning Research, 5:361--397, 2004. Google ScholarDigital Library
- X. Liu, Y. Gong, W. Xu, and S. Zhu. Document clustering with cluster refinement and model selection capabilities. In SIGIR, pages 191--198, 2002. Google ScholarDigital Library
- C. H. Papadimitriou and K. Steiglitz. Combinatorial Optimization : Algorithms and Complexity. Dover Publications, July 1998.Google ScholarDigital Library
- D. Preston, C. E. Brodley, R. Khardon, D. Sulla-Menashe, and M. A. Friedl. Redefining class definitions using constraint-based clustering: an application to remote sensing of the earth's surface. In KDD, pages 823--832, 2010. Google ScholarDigital Library
- B. Scholkopf and A. Smola. Learning with kernels. MIT press Cambridge, Mass, 2002.Google Scholar
- A. Smola, S. Vishwanathan, and Q. Le. Bundle methods for machine learning. NIPS, 20, 2008.Google Scholar
- Q. Tao, D. Chu, and J. Wang. Recursive support vector machines for dimensionality reduction. IEEE Transactions on Neural Networks, 19(1):189--193, 2008. Google ScholarDigital Library
- C. H. Teo, S. V. N. Vishwanathan, A. J. Smola, and Q. V. Le. Bundle methods for regularized risk minimization. Journal of Machine Learning Research, 11:311--365, 2010. Google ScholarDigital Library
- H. Valizadegan and R. Jin. Generalized Maximum Margin Clustering and Unsupervised Kernel Learning. NIPS, 2006.Google Scholar
- A. S. Vishwanathan, A. J. Smola, and S. V. N. Vishwanathan. Kernel methods for missing variables. In AISTAT, pages 325--332, 2005.Google Scholar
- L. Xu, J. Neufeld, B. Larson, and D. Schuurmans. Maximum margin clustering. NIPS, 17:1537--1544, 2005.Google Scholar
- L. Xu and D. Schuurmans. Unsupervised and semi-supervised multi-class support vector machines. In AAAI, pages 904--910, 2005. Google ScholarDigital Library
- T. Yang, R. Jin, A. K. Jain, Y. Zhou, and W. Tong. Unsupervised transfer classification: application to text categorization. In KDD, pages 1159--1168, 2010. Google ScholarDigital Library
- Y. Yang, J. Z. 0003, J. G. Carbonell, and C. Jin. Topic-conditioned novelty detection. In KDD, pages 688--693, 2002. Google ScholarDigital Library
- K. Zhang, I. W. Tsang, and J. T. Kwok. Maximum margin clustering made practical. IEEE Transactions on Neural Networks, 20(4):583--596, 2009. Google ScholarDigital Library
- B. Zhao, F. Wang, and C. Zhang. Efficient maximum margin clustering via cutting plane algorithm. SDM, pages 751--762, 2008.Google ScholarCross Ref
- B. Zhao, F. Wang, and C. Zhang. Efficient multiclass maximum margin clustering. ICML, pages 751--762, 2008.Google ScholarDigital Library
- X. Zhu. Semi-supervised learning literature survey. Computer Science, University of Wisconsin-Madison, 2006.Google Scholar
Index Terms
- Serendipitous learning: learning beyond the predefined label space
Recommendations
Label-activating framework for zero-shot learning
AbstractExisting zero-shot learning (ZSL) models usually learn mappings between visual space and semantic space. However, few of them take the label information into account. Indirect Attribute Prediction (IAP) learns the posterior probability ...
Feature space and label space selection based on Error-correcting output codes for partial label learning
AbstractPartial label learning (PLL) is a type of weakly supervised learning. This paper proposes a new heuristic algorithm for Partial Label learning through the combination of Feature space And Label space using Error-correcting output codes ...
Transductive Maximum Margin Classification of ADHD Using Resting State fMRI
Machine Learning in Medical ImagingAbstractResting-state functional magnetic resonance imaging (rs-fMRI) provides key neural imaging characteristics for quantitative assessment and better understanding of the mechanisms of attention deficit hyperactivity disorder (ADHD). Recent ...
Comments