Outlier detection code for "CURE: Flexible Categorical Data Representation by Hierarchical Coupling Learning" In the paper, we instantiate two instances of CURE framework: CDE and COSH. CDE is used for majority-based tasks,such as clustering. CDE code is in another repository. COSH is used for minority-based tasks, such as outlier detection. COSH is especially effective for high-dimensional data. More information could be found in: https://github.com/jiansonglei.
Citation: @article{jian2018cure, title={CURE: Flexible Categorical Data Representation by Hierarchical Coupling Learning}, author={Jian, Songlei and Cao, Longbing and Lu, Kai and Gao, Hang}, journal={IEEE Transactions on Knowledge and Data Engineering}, year={2018}, publisher={IEEE} }