Dataset Obfuscation: Its Applications to and Impacts on Edge Machine Learning

TL;DR

This paper investigates how dataset obfuscation through noise addition impacts model accuracy, privacy, and model differences in edge machine learning, providing insights into privacy-utility trade-offs and applications.

Contribution

It introduces a comprehensive analysis of dataset obfuscation effects on model performance and privacy, with a novel visualization tool and experimental validation on standard datasets.

Findings

Trade-off between model accuracy and privacy level

Trade-off between model difference and privacy level

Broad applications in edge computing and federated learning

Abstract

Obfuscating a dataset by adding random noises to protect the privacy of sensitive samples in the training dataset is crucial to prevent data leakage to untrusted parties for edge applications. We conduct comprehensive experiments to investigate how the dataset obfuscation can affect the resultant model weights - in terms of the model accuracy, Frobenius-norm (F-norm)-based model distance, and level of data privacy - and discuss the potential applications with the proposed Privacy, Utility, and Distinguishability (PUD)-triangle diagram to visualize the requirement preferences. Our experiments are based on the popular MNIST and CIFAR-10 datasets under both independent and identically distributed (IID) and non-IID settings. Significant results include a trade-off between the model accuracy and privacy level and a trade-off between the model difference and privacy level. The results indicate broad application prospects for training outsourcing in edge computing and guarding against attacks in Federated Learning among edge devices.