Corpus ID: 203626781

Learning Maximally Predictive Prototypes in Multiple Instance Learning

@article{Yksekgnl2019LearningMP,
  title={Learning Maximally Predictive Prototypes in Multiple Instance Learning},
  author={Mert Y{\"u}ksekg{\"o}n{\"u}l and {\"O}zgur Emre Sivrikaya and M. G. Baydogan},
  journal={ArXiv},
  year={2019},
  volume={abs/1910.00965}
}
  • Mert Yüksekgönül, Özgur Emre Sivrikaya, M. G. Baydogan
  • Published 2019
  • Mathematics, Computer Science
  • ArXiv
  • In this work, we propose a simple model that provides permutation invariant maximally predictive prototype generator from a given dataset, which leads to interpretability of the solution and concrete insights to the nature and the solution of a problem. Our aim is to find out prototypes in the feature space to map the collection of instances (i.e. bags) to a distance feature space and simultaneously learn a linear classifier for multiple instance learning (MIL). Our experiments on classical MIL… CONTINUE READING

    Figures, Tables, and Topics from this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 32 REFERENCES
    Attention-based Deep Multiple Instance Learning
    203
    Multiple instance learning with bag dissimilarities
    71
    Combining Instance Information to Classify Bags
    8
    Logistic Regression and Boosting for Labeled Bags of Instances
    217
    Solving the Multiple-Instance Problem: A Lazy Learning Approach
    • Jun Wang, Jean-Daniel Zucker
    • Computer Science
    • 2000
    602
    Deterministic Annealing for Multiple-Instance Learning
    105
    Dissimilarity-Based Ensembles for Multiple Instance Learning
    36
    Multiple-instance learning as a classifier combining problem
    42
    Multi-Instance Kernels
    491