R-SVM+: Robust Learning with Privileged Information

@inproceedings{Li2018RSVMRL,
  title={R-SVM+: Robust Learning with Privileged Information},
  author={Xue Li and Bo Du and Chang Xu and Yipeng Zhang and Lefei Zhang and Dacheng Tao},
  booktitle={IJCAI},
  year={2018}
}
In practice, the circumstance that training and test data are clean is not always satisfied. The performance of existing methods in the learning using privileged information (LUPI) paradigm may be seriously challenged, due to the lack of clear strategies to address potential noises in the data. This paper proposes a novel Robust SVM+ (RSVM+) algorithm based on a rigorous theoretical analysis. Under the SVM+ framework in the LUPI paradigm, we study the lower bound of perturbations of both… 

Figures and Tables from this paper

Retaining Privileged Information for Multi-Task Learning
TLDR
A novel feature matching algorithm is proposed that projects samples from the original feature space and the privilege information space into a joint latent space in a way that informs similarity between training samples and greatly improves the sample efficiency of other related learning tasks.

References

SHOWING 1-10 OF 30 REFERENCES
Fast Algorithms for Linear and Kernel SVM+
TLDR
This paper proposes two efficient algorithms for solving the linear and kernel SVM+, and shows that their new dual problem can be efficiently solved by using the SMO algorithm of the one-class SVM problem.
Learning Using Privileged Information with L-1 Support Vector Machine
  • Lingfeng Niu, Yong Shi, Jianmin Wu
  • Computer Science
    2012 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology
  • 2012
TLDR
This paper introduces the privileged information into the modeling of L-1 support vector machine(SVM) with L-2 SVM with the advantage of spending less time on tuning model parameters and the additional benefits of performing feature selection in the training process.
Object Localization based on Structural SVM using Privileged Information
TLDR
This work tackles object localization problem based on a novel structural SVM using privileged information, where an alternating loss-augmented inference procedure is employed to handle the term in the objective function corresponding to privileged information.
Privileged Multi-label Learning
TLDR
This paper suggests that for each individual label, it cannot only be implicitly connected with other labels via the low-rank constraint over label predictors, but also its performance on examples can receive the explicit comments from other labels together acting as an Oracle teacher.
A new learning paradigm: Learning using privileged information
Transfer Hashing with Privileged Information
TLDR
This work extends the standard learning to hash method, Iterative Quantization (ITQ), in a transfer learning manner, namely ITQ+.
Information Bottleneck Learning Using Privileged Information for Visual Recognition
TLDR
This work establishes an information theoretic principle for leaning any type of visual classifier under this particular setting by extending the information bottleneck method, and by combining it with risk minimization to design a large-margin classifier with an efficient optimization in the primal space.
Person Re-Identification With Metric Learning Using Privileged Information
TLDR
Experimental results on several widely-used data sets demonstrate that the proposed approach is superior to global decision threshold-based methods and outperforms most state-of-the-art results.
Learning to Rank Using Privileged Information
TLDR
This work introduces two maximum-margin techniques that are able to make use of this additional source of information about the training data which however will not be available at test time, and shows that the framework is applicable to several scenarios that have been studied in computer vision before.
...
1
2
3
...