Overcoming Key Weaknesses of Distance-based Neighbourhood Methods using a Data Dependent Dissimilarity Measure

@inproceedings{Ting2016OvercomingKW,
  title={Overcoming Key Weaknesses of Distance-based Neighbourhood Methods using a Data Dependent Dissimilarity Measure},
  author={Kai Ming Ting and Ye Zhu and Mark James Carman and Yue Zhu and Zhi-Hua Zhou},
  booktitle={KDD '16},
  year={2016}
}
  • Kai Ming Ting, Ye Zhu, +2 authors Zhi-Hua Zhou
  • Published in KDD '16 2016
  • Computer Science
  • This paper introduces the first generic version of data dependent dissimilarity and shows that it provides a better closest match than distance measures for three existing algorithms in clustering, anomaly detection and multi-label classification. For each algorithm, we show that by simply replacing the distance measure with the data dependent dissimilarity measure, it overcomes a key weakness of the otherwise unchanged algorithm. 

    Create an AI-powered research feed to stay up to date with new papers like this posted to ArXiv

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 14 CITATIONS

    $\ell m_p$ : A Novel Similarity Measure for Matching Local Image Descriptors

    VIEW 1 EXCERPT
    CITES METHODS

    BISDBx: towards batch-incremental clustering for dynamic datasets using SNN-DBSCAN

    VIEW 3 EXCERPTS
    CITES METHODS
    HIGHLY INFLUENCED

    Clustering graphs using random trees

    VIEW 3 EXCERPTS
    CITES METHODS & BACKGROUND

    Mass estimation as a means to address shortcomings of data depth and density

    VIEW 2 EXCERPTS
    CITES BACKGROUND & METHODS

    Proximity Forest: an effective and scalable distance-based classifier for time series

    VIEW 1 EXCERPT
    CITES METHODS

    A novel structural mass based dissimilarity measure

    VIEW 10 EXCERPTS
    CITES BACKGROUND & METHODS
    HIGHLY INFLUENCED

    A Novel Data Dependent Similarity Measure Algorithm Based on Attribute Selection

    VIEW 3 EXCERPTS
    CITES METHODS

    References

    Publications referenced by this paper.
    SHOWING 1-6 OF 6 REFERENCES

    An information-theoretic definition of similarity

    • D. Lin
    • Proceedings of the Fifteenth International Conference on Machine Learning, pages 296–304, San Francisco, CA, USA,
    • 1998
    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL