Mitigating the ICA Attack against Rotation‐Based Transformation for Privacy Preserving Clustering

The rotation‐based transformation (RBT) for privacy preserving data mining is vulnerable to the independent component analysis (ICA) attack. This paper introduces a modified multiple‐rotation‐based transformation technique for special mining applications, mitigating the ICA attack while maintaining the advantages of the RBT.