Data Leakage

Data leakage in machine learning refers to the unintended exposure of sensitive information from training data through model outputs or intermediate computations. Current research focuses on detecting and mitigating leakage in various contexts, including federated learning, large language models, and recommendation systems, employing techniques like differential privacy and adversarial training to enhance privacy while maintaining model accuracy. Understanding and addressing data leakage is crucial for ensuring the responsible development and deployment of machine learning systems, particularly in sensitive domains like healthcare and finance, and for establishing reliable benchmarks for model evaluation.

Papers

November 21, 2022

Data Leakage and Evaluation Issues in Micro-Expression Analysis
Tuomas Varanka, Yante Li, Wei Peng, Guoying Zhao
Data Leakage Facial Action Unit Micro Expression Unified Evaluation

November 7, 2022

Towards learning to explain with concept bottleneck models: mitigating information leakage
Joshua Lockhart, Nicolas Marchesotti, Daniele Magazzeni, Manuela Veloso
Concept Bottleneck Model Soft Label Data Leakage Model Trustworthiness Concept Prediction Soft Prediction

November 1, 2022

An Empirical Study on Data Leakage and Generalizability of Link Prediction Models for Issues and Commits
Maliheh Izadi, Pooya Rostami Mazrae, Tom Mens, Arie van Deursen
Transfer Learning Empirical Study Link Prediction Stronger Generalizability Training Model Data Leakage Data Splitting Link Analysis

October 6, 2022

CANIFE: Crafting Canaries for Empirical Privacy Measurement in Federated Learning
Samuel Maddock, Alexandre Sablayrolles, Pierre Stock
Differential Privacy Privacy Guarantee Data Leakage Empirical Privacy

October 4, 2022

TabLeak: Tabular Data Leakage in Federated Learning
Mark Vero, Mislav Balunović, Dimitar I. Dimitrov, Martin Vechev
Tabular Datasets Data Leakage Reconstruction Attack FedAvg Algorithm Leakage Attack

September 21, 2022

Measuring and Controlling Split Layer Privacy Leakage Using Fisher Information
Kiwan Maeng, Chuan Guo, Sanjay Kariyappa, Edward Suh
Split Learning Measurement System Privacy Preservation Data Leakage Fisher Information Reconstruction Attack Leakage Attack

August 22, 2022

Split-U-Net: Preventing Data Leakage in Split Learning for Collaborative Multi-Modal Brain Tumor Segmentation
Holger R. Roth, Ali Hatamizadeh, Ziyue Xu, Can Zhao, Wenqi Li, Andriy Myronenko, Daguang Xu
Neural Network Deep Learning Model Medical Image Segmentation Tumor Segmentation Split Learning Data Leakage Biomedical Image Segmentation

June 24, 2022

June 7, 2022

Data Stealing Attack on Medical Images: Is it Safe to Export Networks from Data Lakes?
Huiyu Li, Nicholas Ayache, Hervé Delingette
Medical Image Data Leakage Privacy Preserving Machine Learning Lossy Image Compression Image Codecs Data Lake

May 8, 2022

Private Eye: On the Limits of Textual Screen Peeking via Eyeglass Reflections in Video Conferencing
Yan Long, Chen Yan, Shilin Xiao, Shivan Prasad, Wenyuan Xu, Kevin Fu
Continuum Limit Data Leakage Screen Content Video Conferencing Corneal Reflection

February 21, 2022

Inflation of test accuracy due to data leakage in deep learning-based classification of OCT images
Iulian Emil Tampu, Anders Eklund, Neda Haj-Hosseini
Optical Coherence Tomography Deep Learning Based Test Time Data Leakage

February 14, 2022

Do Gradient Inversion Attacks Make Federated Learning Unsafe?
Ali Hatamizadeh, Hongxu Yin, Pavlo Molchanov, Andriy Myronenko, Wenqi Li, Prerna Dogra, Andrew Feng, Mona G. Flores, Jan Kautz, Daguang Xu, Holger R. Roth
Deep Neural Network Differential Privacy Federated Learning Data Leakage Gradient Inversion Attack

February 7, 2022

Redactor: A Data-centric and Individualized Defense Against Inference Attacks
Geon Heo, Steven Euijong Whang
Data Centric Private Data Decision Boundary Data Leakage Inference Attack Data Redaction

January 21, 2022

TOFU: Towards Obfuscated Federated Updates by Encoding Weight Updates into Gradients from Proxy Data
Isha Garg, Manish Nagaraj, Kaushik Roy
Federated Averaging Data Leakage Weight Update Optimization Proxy Gradient Sharing Federated Data Sanitization Defense

December 6, 2021

When the Curious Abandon Honesty: Federated Learning Is Not Private
Franziska Boenisch, Adam Dziedzic, Roei Schuster, Ali Shahin Shamsabadi, Ilia Shumailov, Nicolas Papernot
Deep Convolutional Neural Network Data Leakage Curiosity Inducing Situation Data Reconstruction Attack Field Trap Gradient Sharing

November 9, 2021

Can Information Flows Suggest Targets for Interventions in Neural Circuits?
Praveen Venkatesh, Sanghamitra Dutta, Neil Mehta, Pulkit Grover
Data Leakage Information Flow Health Intervention Neural Circuit Multiple Target

Data Leakage

Papers

Data Leakage and Evaluation Issues in Micro-Expression Analysis

Towards learning to explain with concept bottleneck models: mitigating information leakage

An Empirical Study on Data Leakage and Generalizability of Link Prediction Models for Issues and Commits

CANIFE: Crafting Canaries for Empirical Privacy Measurement in Federated Learning

TabLeak: Tabular Data Leakage in Federated Learning

Measuring and Controlling Split Layer Privacy Leakage Using Fisher Information

Split-U-Net: Preventing Data Leakage in Split Learning for Collaborative Multi-Modal Brain Tumor Segmentation

Data Leakage in Federated Averaging

FLVoogd: Robust And Privacy Preserving Federated Learning

Data Stealing Attack on Medical Images: Is it Safe to Export Networks from Data Lakes?

Private Eye: On the Limits of Textual Screen Peeking via Eyeglass Reflections in Video Conferencing

Inflation of test accuracy due to data leakage in deep learning-based classification of OCT images

Do Gradient Inversion Attacks Make Federated Learning Unsafe?

Redactor: A Data-centric and Individualized Defense Against Inference Attacks

TOFU: Towards Obfuscated Federated Updates by Encoding Weight Updates into Gradients from Proxy Data

When the Curious Abandon Honesty: Federated Learning Is Not Private

Can Information Flows Suggest Targets for Interventions in Neural Circuits?