A privacy attack where an adversary determines whether a specific data sample was used in the training set by analyzing the model’s outputs.
This is a simple code showing how an overfitted logistic regression model can lead to privacy leakage.
The work is inspired by Gautam Kamath’s ML Lectures.