Searcharxiv⌕ Search

arXiv subjects

Jingchen Hu

Publications and source records attributed to Jingchen Hu.

At least 37 records · Page 2Linked to original sources
arXiv2022

Accuracy Gains from Privacy Amplification Through Sampling for Differential Privacy

Recent research in differential privacy demonstrated that (sub)sampling can amplify the level of protection. For example, for $ε$-differential privacy and simple random sampling with sampling rate $r$, the actual privacy guarantee is approximately $rε$, if a value of $ε$ is used to protect the output from the sample. In this paper, we study whether this amplification effect can be exploited systematically to improve the accuracy of the privatized estimate. Specifically, assuming the agency has information for the full population, we ask under which circumstances accuracy gains could be expected, if the privatized estimate would be computed on a random sample instead of the full population. We find that accuracy gains can be achieved for certain regimes. However, gains can typically only be expected, if the sensitivity of the output with respect to small changes in the database does not depend too strongly on the size of the database. We only focus on algorithms that achieve differential privacy by adding noise to the final output and illustrate the accuracy implications for two commonly used statistics: the mean and the median. We see our research as a first step towards understanding the conditions required for accuracy gains in practice and we hope that these findings will stimulate further research broadening the scope of differential privacy algorithms and outputs considered.

stat.AP↗
arXiv2021

Multiple Imputation and Synthetic Data Generation with the R package NPBayesImputeCat

In many contexts, missing data and disclosure control are ubiquitous and challenging issues. In particular at statistical agencies, the respondent-level data they collect from surveys and censuses can suffer from high rates of missingness. Furthermore, agencies are obliged to protect respondents' privacy when publishing the collected data for public use. The NPBayesImputeCat R package, introduced in this paper, provides routines to i) create multiple imputations for missing data, and ii) create synthetic data for statistical disclosure control, for multivariate categorical data, with or without structural zeros. We describe the Dirichlet process mixture of products of multinomial distributions model used in the package, and illustrate various uses of the package using data samples from the American Community Survey (ACS). We also compare results of the missing data imputation to the mice R package and those of the synthetic data generation to the synthpop R package.

stat.CO↗
arXiv2021

Bayesian Pseudo Posterior Mechanism under Asymptotic Differential Privacy

We propose a Bayesian pseudo posterior mechanism to generate record-level synthetic databases equipped with an $(ε,δ)-$ probabilistic differential privacy (pDP) guarantee, where $δ$ denotes the probability that any observed database exceeds $ε$. The pseudo posterior mechanism employs a data record-indexed, risk-based weight vector with weight values $\in [0, 1]$ that surgically downweight the likelihood contributions for high-risk records for model estimation and the generation of record-level synthetic data for public release. The pseudo posterior synthesizer constructs a weight for each data record using the Lipschitz bound for that record under a log-pseudo likelihood utility function that generalizes the exponential mechanism (EM) used to construct a formally private data generating mechanism. By selecting weights to remove likelihood contributions with non-finite log-likelihood values, we guarantee a finite local privacy guarantee for our pseudo posterior mechanism at every sample size. Our results may be applied to \emph{any} synthesizing model envisioned by the data disseminator in a computationally tractable way that only involves estimation of a pseudo posterior distribution for parameters, $θ$, unlike recent approaches that use naturally-bounded utility functions implemented through the EM. We specify mild conditions that guarantee the asymptotic contraction of $δ$ to $0$ over the space of databases. We illustrate our pseudo posterior mechanism on the sensitive family income variable from the Consumer Expenditure Surveys database published by the U.S. Bureau of Labor Statistics. We show that utility is better preserved in the synthetic data for our pseudo posterior mechanism as compared to the EM, both estimated using the same non-private synthesizer, due to our use of targeted downweighting.

stat.ME↗
arXiv2021

Bayesian Estimation of Attribute and Identification Disclosure Risks in Synthetic Data

The synthetic data approach to data confidentiality has been actively researched on, and for the past decade or so, a good number of high quality work on developing innovative synthesizers, creating appropriate utility measures and risk measures, among others, have been published. Comparing to a large volume of work on synthesizers development and utility measures creation, measuring risks has overall received less attention. This paper focuses on the detailed construction of some Bayesian methods proposed for estimating disclosure risks in synthetic data. In the processes of presenting attribute and identification disclosure risks evaluation methods, we highlight key steps, emphasize Bayesian thinking, illustrate with real application examples, and discuss challenges and future research directions. We hope to give the readers a comprehensive view of the Bayesian estimation procedures, enable synthetic data researchers and producers to use these procedures to evaluate disclosure risks, and encourage more researchers to work in this important growing field.

stat.ME↗
arXiv2021

Identification Risks Evaluation of Partially Synthetic Data with the $\texttt{IdentificationRiskCalculation}$ R Package

We extend a general approach to evaluating identification risk of synthesized variables in partially synthetic data. For multiple continuous synthesized variables, we introduce the use of a radius $r$ in the construction of identification risk probability of each target record, and illustrate with working examples. We create the $\texttt{IdentificationRiskCalculation}$ R package to aid researchers and data disseminators in performing these identification risks evaluation calculations. We demonstrate our methods through the R package with applications to a data sample from the Consumer Expenditure Surveys, and discuss the impacts on risk and data utility of 1) the choice of radius $r$, 2) the choice of synthesized variables, and 3) the choice of number of synthetic datasets. We give recommendations for statistical agencies for synthesizing and evaluating identification risk of continuous variables.

stat.ME↗
arXiv2021

Loss of Regularity of Solutions to Shock Reflection Problems by a Non-symmetric Convex Wedge with Potential Flow Equations

In this paper, we study the problem of shock reflection by a wedge, with the potential flow equation, which is a simplification of the Euler System. In the work of M. Feldman and G. Chen, the existence theory of shock reflection problems with the potential flow equation was established, when the wedge is symmetric w.r.t. the direction of the upstream flow. As a natural extension, we study non-symmetric cases, i.e. when the direction of the upstream flow forms a nonzero angle with the symmetry axis of the wedge. The main idea of investigating the existence of solutions to non-symmetric problems is to study the symmetry of the solution. Then difficulties arise such as free boundaries and degenerate ellipticity, especially when ellipticity degenerates on the free boundary. We developed an integral method to overcome these difficulties. Some estimates near the corner of wedge is also established, as an extension of G.Lieberman's work. We proved that in non-symmetric cases, the ideal Lipschitz solution to the potential flow equation, which we call regular solution, does not exist. This suggests that the potential flow solutions to the non-symmetric shock reflection problem, should have some singularity which is not encountered in symmetric case.

math.AP↗
arXiv2021

Bayesian Estimation of Attribute Disclosure Risks in Synthetic Data with the $\texttt{AttributeRiskCalculation}$ R Package

Synthetic data is a promising approach to privacy protection in many contexts. A Bayesian synthesis model, also known as a synthesizer, simulates synthetic values of sensitive variables from their posterior predictive distributions. The resulting synthetic data can then be released in place of the confidential data. An important evaluation prior to synthetic data release is its level of privacy protection, which is often in the form of disclosure risks evaluation. Attribute disclosure, referring to an intruder correctly inferring the confidential values of synthetic records, is one type of disclosure that is challenging to be computationally evaluated. In this paper, we review and discuss in detail some Bayesian estimation approaches to attribute disclosure risks evaluation, with examples of commonly-used Bayesian synthesizers. We create the $\texttt{AttributeRiskCalculation}$ R package to facilitate its implementation, and demonstrate its functionality with examples of evaluating attribute disclosure risks in synthetic samples of the Consumer Expenditure Surveys.

stat.ME↗
arXiv2021

Risk-Efficient Bayesian Data Synthesis for Privacy Protection

Statistical agencies utilize models to synthesize respondent-level data for release to the public for privacy protection. In this work, we efficiently induce privacy protection into any Bayesian synthesis model by employing a pseudo likelihood that exponentiates each likelihood contribution by an observation record-indexed weight in [0, 1], defined to be inversely proportional to the identification risk for that record. We start with the marginal probability of identification risk for a record, which is composed as the probability that the identity of the record may be disclosed. Our application to the Consumer Expenditure Surveys (CE) of the U.S. Bureau of Labor Statistics demonstrates that the marginally risk-adjusted synthesizer provides an overall improved privacy protection; however, the identification risks actually increase for some moderate-risk records after risk-adjusted pseudo posterior estimation synthesis due to increased isolation after weighting; a phenomenon we label "whack-a-mole". We proceed to construct a weight for each record from a collection of pairwise identification risk probabilities with other records, where each pairwise probability measures the joint probability of re-identification of the pair of records, which mitigates the whack-a-mole issue and produces a more efficient set of synthetic data with lower risk and higher utility for the CE data.

stat.ME↗
arXiv2021

Bayesian Data Synthesis and Disclosure Risk Quantification: An Application to the Consumer Expenditure Surveys

The release of synthetic data generated from a model estimated on the data helps statistical agencies disseminate respondent-level data with high utility and privacy protection. Motivated by the challenge of disseminating sensitive variables containing geographic information in the Consumer Expenditure Surveys (CE) at the U.S. Bureau of Labor Statistics, we propose two non-parametric Bayesian models as data synthesizers for the county identifier of each data record: a Bayesian latent class model and a Bayesian areal model. Both data synthesizers use Dirichlet Process priors to cluster observations of similar characteristics and allow borrowing information across observations. We develop innovative disclosure risks measures to quantify inherent risks in the confidential CE data and how those data risks are ameliorated by our proposed synthesizers. By creating a lower bound and an upper bound of disclosure risks under a minimum and a maximum disclosure risks scenarios respectively, our proposed inherent risks measures provide a range of acceptable disclosure risks for evaluating risks level in the synthetic datasets.

stat.AP↗
arXiv2020

Bayesian Computing in the Undergraduate Statistics Curriculum

Bayesian statistics has gained great momentum since the computational developments of the 1990s. Gradually, advances in Bayesian methodology and software have made Bayesian techniques much more accessible to applied statisticians and, in turn, have potentially transformed Bayesian education at the undergraduate level. This article provides an overview on the various options for implementing Bayesian computational methods motivated to achieve particular learning outcomes. The advantages and disadvantages of each computational method are described based on the authors' experience in using these methods in the classroom. The goal is to present guidance on the choice of computation for the instructors who are introducing Bayesian methods in their undergraduate statistics curriculum.

stat.CO↗
arXiv2020

A Bayesian Statistics Course for Undergraduates: Bayesian Thinking, Computing, and Research

We propose a semester-long Bayesian statistics course for undergraduate students with calculus and probability background. We cultivate students' Bayesian thinking with Bayesian methods applied to real data problems. We leverage modern Bayesian computing techniques not only for implementing Bayesian methods, but also to deepen students' understanding of the methods. Collaborative case studies further enrich students' learning and provide experience to solve open-ended applied problems. The course has an emphasis on undergraduate research, where accessible academic journal articles are read, discussed, and critiqued in class. With increased confidence and familiarity, students take the challenge of reading, implementing, and sometimes extending methods in journal articles for their course projects.

stat.OT↗
arXiv2020

Synthesizing geocodes to facilitate access to detailed geographical information in large scale administrative data

We investigate whether generating synthetic data can be a viable strategy for providing access to detailed geocoding information for external researchers, without compromising the confidentiality of the units included in the database. Our work was motivated by a recent project at the Institute for Employment Research (IAB) in Germany that linked exact geocodes to the Integrated Employment Biographies (IEB), a large administrative database containing several million records. We evaluate the performance of three synthesizers regarding the trade-off between preserving analytical validity and limiting disclosure risks: One synthesizer employs Dirichlet Process mixtures of products of multinomials (DPMPM), while the other two use different versions of Classification and Regression Trees (CART). In terms of preserving analytical validity, our proposed synthesis strategy for geocodes based on categorical CART models outperforms the other two. If the risks of the synthetic data generated by the categorical CART synthesizer are deemed too high, we demonstrate that synthesizing additional variables is the preferred strategy to address the risk-utility trade-off in practice, compared to limiting the size of the regression trees or relying on the strategy of providing geographical information only on an aggregated level. We also propose strategies for making the synthesizers scalable for large files, present analytical validity measures and disclosure risk measures for the generated data, and provide general recommendations for statistical agencies considering the synthetic data approach for disseminating detailed geographical information.

stat.AP↗
arXiv2020

Two-Phase Data Synthesis for Income: An Application to the NHIS

We propose a two-phase synthesis process for synthesizing income, a sensitive variable which is usually highly-skewed and has a number of reported zeros. We consider two forms of a continuous income variable: a binary form, which is modeled and synthesized in phase 1; and a non-negative continuous form, which is modeled and synthesized in phase 2. Bayesian synthesis models are proposed for the two-phase synthesis process, and other synthesis models are readily implementable. We demonstrate our methods with applications to a sample from the National Health Interview Survey (NHIS). Utility and risk profiles of generated synthetic datasets are evaluated and compared to results from a single-phase synthesis process.

stat.AP↗
arXiv2020

Bayesian Pseudo Posterior Synthesis for Data Privacy Protection

Statistical agencies utilize models to synthesize respondent-level data for release to the general public as an alternative to the actual data records. A Bayesian model synthesizer encodes privacy protection by employing a hierarchical prior construction that induces smoothing of the real data distribution. Synthetic respondent-level data records are often preferred to summary data tables due to the many possible uses by researchers and data analysts. Agencies balance a trade-off between utility of the synthetic data versus disclosure risks and hold a specific target threshold for disclosure risk before releasing synthetic datasets. We introduce a pseudo posterior likelihood that exponentiates each contribution by an observation record-indexed weight in (0, 1), defined to be inversely proportional to the disclosure risk for that record in the synthetic data. Our use of a vector of weights allows more precise downweighting of high risk records in a fashion that better preserves utility as compared with using a scalar weight. We illustrate our method with a simulation study and an application to the Consumer Expenditure Survey of the U.S. Bureau of Labor Statistics. We demonstrate how the frequentist consistency and uncertainty quantification are affected by the inverse risk-weighting.

math.ST↗
arXiv2020

Loss of Regularity of Solutions of the Lighthill Problem for Shock Diffraction for Potential Flow

We are concerned with the suitability of the main models of compressible fluid dynamics for the Lighthill problem for shock diffraction by a convex corned wedge, by studying the regularity of solutions of the problem, which can be formulated as a free boundary problem. In this paper, we prove that there is no regular solution that is subsonic up to the wedge corner for potential flow. This indicates that, if the solution is subsonic at the wedge corner, at least a characteristic discontinuity (vortex sheet or entropy wave) is expected to be generated, which is consistent with the experimental and computational results. Therefore, the potential flow equation is not suitable for the Lighthill problem so that the compressible Euler system must be considered. In order to achieve the non-existence result, a weak maximum principle for the solution is established, and several other mathematical techniques are developed. The methods and techniques developed here are also useful to the other problems with similar difficulties.

math.AP↗
arXiv2020

Online Statistics Teaching and Learning

For statistics courses at all levels, teaching and learning online poses challenges in different aspects. Particular online challenges include how to effectively and interactively conduct exploratory data analyses, how to incorporate statistical programming, how to include individual or team projects, and how to present mathematical derivations efficiently and effectively. This article draws from the authors' experience with seven different online statistics courses to address some of the aforementioned challenges. One course is an online exploratory data analysis course taught at Bowling Green State University. A second course is an upper level Bayesian statistics course taught at Vassar College and shared among 10 liberal arts colleges through a hybrid model. We alo describes a five-course MOOC specialization on Coursera, offered by Duke University.

stat.OT↗