Noise-Aware Differentially Private Regression via Meta-Learning

1 code implementation • 12 Jun 2024 • Ossi Räisä, Stratis Markou, Matthew Ashman, Wessel P. Bruinsma, Marlon Tobaben, Antti Honkela, Richard E. Turner

One approach to mitigating this issue is pre-training models on simulated data before DP learning on the private data.

Meta-Learning regression

Subsampling is not Magic: Why Large Batch Sizes Work for Differentially Private Stochastic Optimisation

1 code implementation • 6 Feb 2024 • Ossi Räisä, Joonas Jälkö, Antti Honkela

The remaining subsampling-induced variance decreases with larger batch sizes, so large batches reduce the effective total gradient variance.

A Bias-Variance Decomposition for Ensembles over Multiple Synthetic Datasets

1 code implementation • 6 Feb 2024 • Ossi Räisä, Antti Honkela

Recent studies have highlighted the benefits of generating multiple synthetic datasets for supervised learning, from increased accuracy to more effective model selection and uncertainty estimation.

Model Selection

Noise-Aware Statistical Inference with Differentially Private Synthetic Data

2 code implementations • 28 May 2022 • Ossi Räisä, Joonas Jälkö, Samuel Kaski, Antti Honkela

For example, confidence intervals become too narrow, which we demonstrate with a simple experiment.

Imputation Synthetic Data Generation +1