| Popis: |
The modern developments have made our lives’ more digital. As they make our day-to-day activities simpler, they are in wide usage leading to accumulation of huge amounts of data. The analysis of such data will provide useful results for further advancements which can be obtained by the concept of Data Mining. These newly emerging digital exercises involve sharing of personal information to unknown sources, questioning the privacy of the individual. To avoid such situations, techniques of Privacy-Preserving Data Mining involving various data distortions are used. One such technique is Privacy-Preserving Data Classification, balancing both privacy and utility in classification aspects. In this work, the performance analysis of Classification using Support Vector Machine (SVM) on data sets that are heterogeneously distorted is evaluated. |
