Skip to main content
arXiv is now an independent nonprofit! Learn more

Showing 1–7 of 7 results for author: Marchant, N G

Searching in archive stat. Search in all archives.
.
  1. arXiv:2512.05931  [pdf, ps, other

    cs.LG stat.ML

    On the Bayes Inconsistency of Disagreement Discrepancy Surrogates

    Authors: Neil G. Marchant, Andrew C. Cullen, Feng Liu, Sarah M. Erfani

    Abstract: Deep neural networks often fail when deployed in real-world contexts due to distribution shift, a critical barrier to building safe and reliable systems. An emerging approach to address this problem relies on \emph{disagreement discrepancy} -- a measure of how the disagreement between two models changes under a shifting distribution. The process of maximizing this measure has seen applications in… ▽ More

    Submitted 5 December, 2025; originally announced December 2025.

    Comments: 37 pages, 7 figures

  2. arXiv:2405.13375  [pdf, ps, other

    cs.LG stat.ML

    Adaptive Data Analysis for Growing Data

    Authors: Neil G. Marchant, Benjamin I. P. Rubinstein

    Abstract: Reuse of data in adaptive workflows poses challenges regarding overfitting and the statistical validity of results. Previous work has demonstrated that interacting with data via differentially private algorithms can mitigate overfitting, achieving worst-case generalization guarantees with asymptotically optimal data requirements. However, such past work assumes data is static and cannot accommodat… ▽ More

    Submitted 12 November, 2025; v1 submitted 22 May, 2024; originally announced May 2024.

    Comments: 28 pages, 6 figures, camera ready version for NeurIPS 2025. Addresses a bug in the computation of the JLNRSS and JLNRSS+ curves in Fig 2 of the previous version. Generalizes the results to non-uniform differential privacy

  3. arXiv:2302.01757  [pdf, other

    cs.CR cs.LG stat.ML

    RS-Del: Edit Distance Robustness Certificates for Sequence Classifiers via Randomized Deletion

    Authors: Zhuoqun Huang, Neil G. Marchant, Keane Lucas, Lujo Bauer, Olga Ohrimenko, Benjamin I. P. Rubinstein

    Abstract: Randomized smoothing is a leading approach for constructing classifiers that are certifiably robust against adversarial examples. Existing work on randomized smoothing has focused on classifiers with continuous inputs, such as images, where $\ell_p$-norm bounded adversaries are commonly studied. However, there has been limited work for classifiers with discrete or variable-size inputs, such as for… ▽ More

    Submitted 24 January, 2024; v1 submitted 30 January, 2023; originally announced February 2023.

    Comments: Final camera-ready version for NeurIPS 2023. 36 pages, 7 figures, 12 tables. Includes 20 pages of appendices. Code available at https://github.com/Dovermore/randomized-deletion

  4. Bayesian Graphical Entity Resolution Using Exchangeable Random Partition Priors

    Authors: Neil G. Marchant, Benjamin I. P. Rubinstein, Rebecca C. Steorts

    Abstract: Entity resolution (record linkage or deduplication) is the process of identifying and linking duplicate records in databases. In this paper, we propose a Bayesian graphical approach for entity resolution that links records to latent entities, where the prior representation on the linkage structure is exchangeable. First, we adopt a flexible and tractable set of priors for the linkage structure, wh… ▽ More

    Submitted 7 January, 2023; originally announced January 2023.

    Comments: 27 pages, 4 figures, 3 tables. Includes 37 pages of appendices. This is an accepted manuscript to be published in the Journal of Survey Statistics and Methodology

  5. arXiv:2006.06963  [pdf, other

    cs.LG cs.IR stat.ML

    Needle in a Haystack: Label-Efficient Evaluation under Extreme Class Imbalance

    Authors: Neil G. Marchant, Benjamin I. P. Rubinstein

    Abstract: Important tasks like record linkage and extreme classification demonstrate extreme class imbalance, with 1 minority instance to every 1 million or more majority instances. Obtaining a sufficient sample of all classes, even just to achieve statistically-significant evaluation, is so challenging that most current approaches yield poor estimates or incur impractical cost. Where importance sampling ha… ▽ More

    Submitted 2 June, 2021; v1 submitted 12 June, 2020; originally announced June 2020.

    Comments: 30 pages, 8 figures, updated to match version accepted for publication at KDD'21

    ACM Class: H.3.4; I.5.2

  6. arXiv:1909.06039  [pdf, other

    stat.CO cs.DB cs.LG stat.ML

    d-blink: Distributed End-to-End Bayesian Entity Resolution

    Authors: Neil G. Marchant, Andee Kaplan, Daniel N. Elazar, Benjamin I. P. Rubinstein, Rebecca C. Steorts

    Abstract: Entity resolution (ER; also known as record linkage or de-duplication) is the process of merging noisy databases, often in the absence of unique identifiers. A major advancement in ER methodology has been the application of Bayesian generative models, which provide a natural framework for inferring latent entities with rigorous quantification of uncertainty. Despite these advantages, existing mode… ▽ More

    Submitted 22 September, 2020; v1 submitted 13 September, 2019; originally announced September 2019.

    Comments: 32 pages, 6 figures, 5 tables. Includes 22 pages of supplementary material. This revision incorporates a case study on the 2010 U.S. Decennial Census

    MSC Class: 62F15; 65C40; 68W15

  7. arXiv:1703.00617  [pdf, other

    cs.LG cs.DB stat.ML

    In Search of an Entity Resolution OASIS: Optimal Asymptotic Sequential Importance Sampling

    Authors: Neil G. Marchant, Benjamin I. P. Rubinstein

    Abstract: Entity resolution (ER) presents unique challenges for evaluation methodology. While crowdsourcing platforms acquire ground truth, sound approaches to sampling must drive labelling efforts. In ER, extreme class imbalance between matching and non-matching records can lead to enormous labelling requirements when seeking statistically consistent estimates for rigorous evaluation. This paper addresses… ▽ More

    Submitted 25 June, 2017; v1 submitted 1 March, 2017; originally announced March 2017.

    Comments: 13 pages, 5 figures