Faster Algorithms for User-Level Private Stochastic Convex Optimization

Authors: Andrew Lowy, Daogao Liu, Hilal Asi

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical This is a theoretical paper without experiments.
Researcher Affiliation Collaboration Andrew Lowy Wisconsin Institute for Discovery University of Wisconsin-Madison alowy@wisc.edu; Daogao Liu Department of Computer Science University of Washington liudaogao@gmail.com; Hilal Asi Apple Machine Learning Research hilal.asi94@gmail.com
Pseudocode Yes Algorithm 1: User-Level DP Phased SGD with Outlier Iterate Removal and Output Perturbation; Algorithm 2: User-Level DP Accelerated Minibatch SGD( b Fi, Ti, Ki, xi 1, τ, ε, δ); Algorithm 3: User-Level DP Accelerated Phased ERM with Outlier Gradient Removal
Open Source Code No This is a theoretical paper without experiments. The paper does not provide any statement about releasing source code for the described methodology.
Open Datasets No This is a theoretical paper without experiments. The paper does not perform experiments on specific datasets.
Dataset Splits No This is a theoretical paper without experiments. The paper does not perform experiments, hence no dataset splits are provided.
Hardware Specification No This is a theoretical paper without experiments. The paper does not describe hardware used for experiments.
Software Dependencies No This is a theoretical paper without experiments. The paper does not list specific software dependencies with version numbers for experimental replication.
Experiment Setup No This is a theoretical paper without experiments. The paper does not provide specific experimental setup details such as hyperparameter values.