Label Inference Attacks in Federated Learning

Description

• Explored label inference attacks in vertical federated learning (VFL) to evaluate the risks of revealing label information during training.

Tech Stack

• Python, PyTorch, NumPy, Scikit-learn.

Contributions

• Developed and optimized a label inference attack algorithm in PyTorch.
• Preprocessed and analyzed data to ensure compatibility with federated learning frameworks.

Supervisor:

Professor Oshani Seneviratne