Research map / Federated learning and privacy / Differential privacy
Online and Differentially-Private Tensor Decomposition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use random initialization on sphere with improved noise analysis (noise tolerance ||Delta||_op = O(lambda_min/sqrt(d)) for round noise) instead of O(lambda_min/d) | Robust tensor power method · Theory (round noise) | noise tolerance | Helped |
| Streaming (online) tensor power method with data-association updates instead of forming the full tensor | Robust tensor power method · Theory (streaming, sub-Gaussian noise) | memory and sample complexity | Helped |
| Noise calibration (Gaussian noise added at every power iteration) for differential privacy | Robust tensor power method · Theory (DP, incoherent basis) | privacy parameter epsilon requirement | Helped |
| Input perturbation (adding Gaussian tensor before power iterations) for differential privacy instead of per-iteration noise calibration | Robust tensor power method · Theory (DP) | required privacy parameter epsilon | Hurt |
| Whitening via marginalizing one view with random theta and matrix SVD instead of the improved tensor power analysis | Tensor decomposition baseline · Theory (whitening comparison) | noise tolerance bound | Hurt |
Closest papers by meaning
- Low-Rank Tucker Decomposition of Large Tensors Using TensorSketch (NeurIPS 2018 · similarity 0.87)
- Landscape analysis of an improved power method for tensor decomposition (NeurIPS 2021 · similarity 0.87)
- Expected Tensor Decomposition with Stochastic Gradient Descent (AAAI 2016 · similarity 0.86)
- Settling Pure Differentially Private Covariance Estimation (NeurIPS 2026 · similarity 0.86)
- Perturb-and-Project: Differentially Private Similarities and Marginals (ICML 2024 · similarity 0.85)
- Distribution-Aware Tensor Decomposition for Compression of Convolutional Neural Networks (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.