diff options
Diffstat (limited to 'synthetic/bck/introduction.tex')
| -rw-r--r-- | synthetic/bck/introduction.tex | 9 |
1 files changed, 9 insertions, 0 deletions
diff --git a/synthetic/bck/introduction.tex b/synthetic/bck/introduction.tex new file mode 100644 index 0000000..e8969c2 --- /dev/null +++ b/synthetic/bck/introduction.tex @@ -0,0 +1,9 @@ +\subsection{Research questions} +\label{sec:question} + +%\textbf{What is the impact of using synthetic data instead of real data on users' privacy when training machine learning models?} +\textbf{How does using synthetic data instead of real data affect users' privacy in the context of training machine learning models?} + +User's privacy and neural network clash at two levels: membership inference and attribute inference. +Membership inference refers to the possibility of infering weather or not a data record belongs to the training data. +Attribute inference refers to how a trained model can be leveraged to infer a sensitive attribute such as the race or the gender. |