Skip to main content
Search

# Membership Inference Attack

3 related articles

Subscribe to updates
  1. arXiv

    Continually updated models leak when data was used: TMIA reads confidence trajectories to time sample entry and exit, with timestamp F1 above 83%

    Continually updated models keep a sequence of queryable snapshots, so an attacker can infer not only whether a sample was used in training but when it entered and left. TMIA treats a sample's confidence across snapshots as a time series, labels entry and exit points with a sliding-window bidirectional LSTM, and pairs them chronologically into valid membership intervals. Across three language models and three image classifiers, TMIA reaches timestamp-level F1 above 83% in every language-model setting and improves interval-level F1 over both baselines by at least 21.82%.

  2. Four network-traffic generators leak their training data: membership inference reaches 88% TPR and network identifiers can be fully recovered

    The study introduces a privacy measurement suite for synthetic network traffic spanning membership inference, data extraction, and network-specific attacks on identifiers, attributes, and topology, and evaluates four representative generators (NetShare, NetDiffusion, TrafficLLM, NetSSM) across five datasets, finding that even with black-box-only attackers and minimal training, membership inference reaches up to 0.88 TPR at FPR≤0.01 and up to 100% of network identifiers can be recovered, while anonymization and DP noise each cover only part of the risk at a utility cost.

  3. A PCA-plus-local-differential-privacy framework for agricultural data sharing trades about 5.33% average accuracy loss for privacy on real datasets

    The work proposes a privacy-preserving data-sharing and collaboration framework for digital agriculture that combines Principal Component Analysis (PCA) dimensionality reduction with Laplacian-noise-based local differential privacy (ε-LDP), aggregates farmer data in a sandbox environment, and uses K-Means clustering and nearest-neighbor algorithms to recommend potential collaborators, enabling researchers to train personalized models via federated learning or directly on aggregated privacy-protected data; validated on the Wisconsin Farmer's Market and Crop Recommendation real-world datasets, model accuracy shows an average loss of about 5.33% compared with centralized raw-data training, with robustness against attacks such as membership inference assessed via power analysis.

You’re all caught up