Abstract:Author(s): Song, Shuang | Advisor(s): Chaudhuri, Kamalika | Abstract: Modern machine learning increasingly involves personal data, such as healthcare, financial and user behavioral information. However, models trained on such data can reveal detailed information about the data and cause a serious privacy breach. Consequently, it is important to design algorithms that can analyze the sensitive data while still preserving privacy. This thesis advances the state-of-the-art of privacy-preserving machine learning in the following two major aspects.First, this thesis addresses the challenges in differentially private large-scale machine learning. On the one hand, with a large amount of sensitive user data, privacy-preserving learning algorithms are expected to achieve improved utility. On the other hand, big data imposes additional challenges, including performance (a.k.a Volume), data noise (a.k.a Veracity), a large number of classes and distributed sources (a.k.a Variety). This thesis presents (1) private versions of the widely used stochastic gradient descent (SGD) algorithm with generalization to data from multiple sources of different privacy requirements, and (2) an improved version of the Private Aggregation of Teacher Ensembles (PATE) framework which can scale to learning tasks with a large number of output classes and uncurated, imbalanced training data.Second, this thesis considers privacy-preserving data analysis beyond tabular data. Differential privacy is best suited for tabular data, where each record corresponds to all the information about an individual and records are independent of each other. However, many real-world applications involve non-tabular sensitive data, such as epidemic transmission graphs and measurement of the physical activity of a single subject across time. To analyze disease transmission graphs, or more general, graphs with sensitive information stored in each node, this thesis considers privacy-preserving continual release of graph statistics, such as the percentages of highly-active patients over time. The proposed algorithm outperforms the baselines in utility over a range of parameters. For physical activity measurement, or more generally, data with correlation, this thesis looks at a recent generalization of differential privacy, called Pufferfish privacy, that addresses privacy concerns in correlated data. Two mechanisms that work under different scenarios are proposed, and one of them is evaluated on real and synthetic time-series data.

Privacy-preserving data release leveraging optimal transport and particle gradient descent

PrivSyn: Differentially Private Data Synthesis

Differentially Private Synthetic Data with Private Density Estimation

Generating tabular datasets under differential privacy

Assessment of differentially private synthetic data for utility and fairness in end-to-end machine learning pipelines for tabular data

DP-CDA: An Algorithm for Enhanced Privacy Preservation in Dataset Synthesis Through Randomized Mixing

Approximate, Adapt, Anonymize (3A): a Framework for Privacy Preserving Training Data Release for Machine Learning

Generating Private Synthetic Data with Genetic Algorithms

Differentially Private Synthetic High-dimensional Tabular Stream

A Data Synthesis Approach Based on Local Differential Privacy

Differentially Private Synthetic Data Generation via Lipschitz-Regularised Variational Autoencoders

Quantifying and Mitigating Privacy Risks for Tabular Generative Models

Differentially Private Synthetic Data Using KD-Trees

Differentially Private Synthetic Heavy-tailed Data

Privacy-Preserving Algorithms for Machine Learning

Differentially Private Synthetic Data: Applied Evaluations and Enhancements

pMSE Mechanism: Differentially Private Synthetic Data with Maximal Distributional Similarity

Training generative models from privatized data

Selecting Optimal Subset to release under Differentially Private M-estimators from Hybrid Datasets

Stochastic gradient descent with differentially private updates