Feature Selection on Anonymized Datasets.

Privacy concerns have become increasingly prominent in data analysis, leading to adoption of privacy-preserving approaches including anonymization techniques. However, the application of anonymization techniques may impact the significance of features in the anonymized dataset, potentially affecting the accuracy and reliability of subsequent machine learning tasks. In this paper, we present a comprehensive comparative study to investigate the preservation of feature significance in the original dataset versus its anonymized version. We conduct a set of experiments to provide valuable insights into the implications of anonymization techniques on preserving the features' relevancy, which advances privacy-aware data analysis in various domains.