27 citations · 110 across the 24 of their papers we have counts for
28 papers
Not Every Time and Frequency Need to Be Forgotten in Diffusion Unlearning
Jinseong Park, Mijung Park
Data unlearning aims to remove the influence of specific training samples from a trained model. In fine-tuning methods, data unlearning relies primarily on loss maximization over f…
Optimizable Model Averaging for Ensembling Heterogeneous Model Outputs
Mijung Park
Foundation model ensembles can improve accuracy and robustness, but full fine-tuning and parameter-space merging are often expensive or impossible when candidate models have differ…
Pre-Pruning and Gradient-Dropping Improve Differentially Private Image Classification
Kamil Adamczewski, Yingchen He, Mijung Park
Scalability is a significant challenge when it comes to applying differential privacy to training deep neural networks. The commonly used DP-SGD algorithm struggles to maintain a h…
DP-LDMs: Differentially Private Latent Diffusion Models
Michael F. Liu, Saiyue Lyu, Margarita Vinaroz +1
Diffusion models (DMs) are one of the most widely used generative models for producing high quality images. However, a flurry of recent papers points out that DMs are least private…
Differential Privacy Meets Neural Network Pruning
Kamil Adamczewski, Mijung Park
A major challenge in applying differential privacy to training deep neural network models is scalability.The widely-used training algorithm, differentially private stochastic gradi…
Differentially Private Neural Tangent Kernels for Privacy-Preserving Data Generation
Yilin Yang, Kamil Adamczewski, Danica J. Sutherland +2
Maximum mean discrepancy (MMD) is a particularly useful distance metric for differentially private data generation: when used with finite-dimensional features it allows us to summa…
Differentially Private Kernel Inducing Points using features from ScatterNets (DP-KIP-ScatterNet) for Privacy Preserving Data Distillation
Margarita Vinaroz, Mi Jung Park
Data distillation aims to generate a small data set that closely mimics the performance of a given learning algorithm on the original data set. The distilled dataset is hence usefu…
Pre-trained Perceptual Features Improve Differentially Private Image Generation
Fredrik Harder, Milad Jalali Asadabadi, Danica J. Sutherland +1
Training even moderately-sized generative models with differentially-private stochastic gradient descent (DP-SGD) is difficult: the required level of noise for reasonable levels of…
Differentially private stochastic expectation propagation (DP-SEP)
Margarita Vinaroz, Mijung Park
We are interested in privatizing an approximate posterior inference algorithm called Expectation Propagation (EP). EP approximates the posterior by iteratively refining approximati…
Hermite Polynomial Features for Private Data Generation
Margarita Vinaroz, Mohammad-Amin Charusaie, Frederik Harder +2
Kernel mean embedding is a useful tool to represent and compare probability measures. Despite its usefulness, kernel mean embedding considers infinite-dimensional features, which a…
Dirichlet Pruning for Neural Network Compression
Kamil Adamczewski, Mijung Park
We introduce Dirichlet pruning, a novel post-processing technique to transform a large neural network model into a compressed one. Dirichlet pruning is a form of structured pruning…
Bayesian Importance of Features (BIF)
Kamil Adamczewski, Frederik Harder, Mijung Park
We introduce a simple and intuitive framework that provides quantitative explanations of statistical models through the probabilistic assessment of input feature importance. The co…
DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-Preserving Data Generation
Frederik Harder, Kamil Adamczewski, Mijung Park
We propose a differentially private data generation paradigm using random feature representations of kernel mean embeddings when comparing the distribution of true data with that o…
DP-MAC: The Differentially Private Method of Auxiliary Coordinates for Deep Learning
Frederik Harder, Jonas Köhler, Max Welling +1
Developing a differentially private deep learning algorithm is challenging, due to the difficulty in analyzing the sensitivity of objective functions that are typically used to tra…
ABCDP: Approximate Bayesian Computation with Differential Privacy
Mijung Park, Margarita Vinaroz, Wittawat Jitkrittum
We develop a novel approximate Bayesian computation (ABC) framework, ABCDP, that produces differentially private (DP) and approximate posterior samples. Our framework takes advanta…
Neuron ranking -- an informed way to condense convolutional neural networks architecture
Kamil Adamczewski, Mijung Park
Convolutional neural networks (CNNs) in recent years have made a dramatic impact in science, technology and industry, yet the theoretical mechanism of CNN architecture design remai…
Interpretable and Differentially Private Predictions
Frederik Harder, Matthias Bauer, Mijung Park
Interpretable predictions, where it is clear why a machine learning model has made a particular decision, can compromise privacy by revealing the characteristics of individual data…
Privacy-Preserving Causal Inference via Inverse Probability Weighting
Si Kai Lee, Luigi Gresele, Mijung Park +1
The use of inverse probability weighting (IPW) methods to estimate the causal effect of treatments from observational studies is widespread in econometrics, medicine and social sci…
Radial and Directional Posteriors for Bayesian Neural Networks
Changyong Oh, Kamil Adamczewski, Mijung Park
We propose a new variational family for Bayesian neural networks. We decompose the variational posterior into two components, where the radial component captures the strength of ea…
A Differentially Private Kernel Two-Sample Test
Anant Raj, Ho Chung Leon Law, Dino Sejdinovic +1
Kernel two-sample testing is a useful statistical tool in determining whether data samples arise from different distributions without imposing any parametric assumptions on those d…
Variational Bayes In Private Settings (VIPS)
Mijung Park, James Foulds, Kamalika Chaudhuri +1
Many applications of Bayesian data analysis involve sensitive information, motivating methods which ensure that privacy is protected. We introduce a general privacy-preserving fram…
Private Topic Modeling
Mijung Park, James Foulds, Kamalika Chaudhuri +1
We develop a privatised stochastic variational inference method for Latent Dirichlet Allocation (LDA). The iterative nature of stochastic variational inference presents challenges:…
A note on privacy preserving iteratively reweighted least squares
Mijung Park, Max Welling
Iteratively reweighted least squares (IRLS) is a widely-used method in machine learning to estimate the parameters in the generalised linear models. In particular, IRLS for L1 mini…
DP-EM: Differentially Private Expectation Maximization
Mijung Park, Jimmy Foulds, Kamalika Chaudhuri +1
The iterative nature of the expectation maximization (EM) algorithm presents a challenge for privacy-preserving estimation, as each iteration increases the amount of noise needed.…