High Dimensional Differentially Private Stochastic Optimization with Heavy-tailed Data
Summary: First study of high-dimensional DP-SCO under heavy-tailed data: derives high-probability excess-risk bounds for polytope constraints and an improved LASSO rate O~((log d)/(n ε)^2) under bounded fourth moments. Introduces truncated DP-HT for sparse learning achieving O~(s^2 log d/(n ε)) and an l0-constrained method with O~(s^{3/2} log d/(n ε)), nearly optimal up to √s. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Lijie Hu (King Abdullah University of Science and Technology)
- 2. Shuo Ni (University of Southern California)
- 3. Hanshen Xiao (Massachusetts Institute of Technology)
- 4. Di Wang (King Abdullah University of Science and Technology)
BibTeX Citation
@inproceedings{hu_pods22,
address = {New York, NY, USA},
series = {{PODS} '22},
title = {{High Dimensional Differentially Private Stochastic Optimization with Heavy-tailed Data}},
url = {https://dl.acm.org/doi/10.1145/3517804.3524144},
doi = {10.1145/3517804.3524144},
booktitle = {Proceedings of the {ACM} {SIGMOD} Symposium on {Principles} of {Database} {Systems}},
publisher = {Association for Computing Machinery},
author = {Hu, Lijie and Ni, Shuo and Xiao, Hanshen and Wang, Di},
year = {2022}
}
Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 0 of 0 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 4,799 | Scalable Asynchronous Gradient Descent Optimization for Out-of-Core Models | 2017 | VLDB |
| 2 | 10,162 | On Sketching Trimmed Statistics | 2026 | PODS |
| 3 | 611 | Optimizing Linear Counting Queries Under Differential Privacy | 2010 | PODS |
| 4 | 7,841 | Quantifying identifiability to choose and audit epsilon in differentially private deep learning | 2021 | VLDB |
| 5 | 7,898 | A workload-adaptive mechanism for linear queries under local differential privacy | 2020 | VLDB |
| 6 | 5,314 | Optimal Differentially Private Algorithms for k-Means Clustering | 2018 | PODS |
| 7 | 8,929 | Distributed and Streaming Linear Programming in Low Dimensions | 2019 | PODS |
| 8 | 4,562 | Private Multiplicative Weights Beyond Linear Queries | 2015 | PODS |
| 9 | 10,484 | Privacy Loss of Noise Perturbation via Concentration Analysis of A Product Measure | 2026 | SIGMOD |
| 10 | 7,517 | Private Incremental Regression | 2017 | PODS |