BayesianGDPR

Objectives

The General Data Protection Regulation (GDPR) states data should be processed lawfully, fairly and transparently. With this in mind, BayesianGDPR project aims to integrate the legal non-discriminatory principles of GDPR into automated machine-learning systems in a transparent manner. It will do so by using a novel Bayesian approach to model all sources of uncertainty, and taking into account feedback from humans and future consequences of their outputs. BayesianGDPR will provide organisations that rely on machine learning technologies with concrete tools allowing them compliance with the non-discriminatory principles of GDPR and similar laws. The project’s achievements will have an impact on computational law research and its integration into mainstream legal practice. It will also promote public confidence in machine learning systems.

Principal Investigator

• Novi Quadrianto

Team Members

• Leonidas Gee (PhD Student 2023-present)
• Myles Bartlett (Postdoc 2023-present; PhD Student, 2022)
• Sara Romiti (PhD Student, 2022)
• Yeat Jeng Ng (PhD Student 2020–present)
• Ainhize Barrainkua (BCAM PhD Student 2021–present)
• Gergely Dániel Németh (ELLIS PhD student 2021–present)
• Jonas Klesen (ELLIS PhD student 2021–present)
• Thomas Kehrenberg (Postdoc 2021-present)
• Miri Zilka (Postdoc 2020–2021)
• Oliver Thomas (PhD Student 2021)

Publications

• Ainhize Barrainkua, Paula Gordaliza, Jose A. Lozano, Novi Quadrianto. Uncertainty Matters: Stable Conclusions under Unstable Assessment of Fairness Results. International Conference on Artificial Intelligence and Statistics AISTATS, Valencia, Spain, 2024.
• Gergely D. Németh, Miguel Angel Lozano, Novi Quadrianto, Nuria Oliver. Addressing Membership Inference Attack in Federated Learning with Model Compression. arXiv, 2023.
• Leonidas Gee, Andrea Zugarini, Novi Quadrianto. Are Compressed Language Models Less Subgroup Robust?. Conference on Empirical Methods in Natural Language Processing EMNLP, Singapore, 2023.
• Ainhize Barrainkua, Paula Gordaliza, Jose A. Lozano, Novi Quadrianto. Preserving the Fairness Guarantees of Classifiers in Changing Environments: a Survey. ACM Computing Surveys, 2023.
• Myles Bartlett, Sara Romiti, Viktoriia Sharmanska, Novi Quadrianto. Okapi: Generalising Better by Making Statistical Matches Match. Thirty-Sixth Conference on Neural Information Processing Systems NeurIPS, New Orleans, Louisiana, USA, 2022.
• Sara Romiti, Christopher Inskip, Viktoriia Sharmanska, Novi Quadrianto. RealPatch: A Statistical Matching Framework for Model Patching with Real Samples. European Conference on Computer Vision ECCV, Tel-Aviv, Israel, 2022.
• Thomas Kehrenberg, Myles Bartlett, Viktoriia Sharmanska, Novi Quadrianto. Addressing Missing Sources with Adversarial Support-Matching. arXiv, 2022.
• Gergely D. Németh, Miguel Angel Lozano, Novi Quadrianto, Nuria Oliver. A Snapshot of the Frontiers of Client Selection in Federated Learning. Transactions on Machine Learning Research TMLR, 2022.
• Oliver Thomas, Miri Zilka, Adrian Weller, Novi Quadrianto. An Algorithmic Framework for Positive Action. ACM conference on Equity and Access in Algorithms, Mechanisms, and Optimization EAAMO, Virtual, 2021 [selected for oral presentation].
• Bradley Butcher, Vincent Huang, Christopher Robinson, Jeremy Reffin, Sema Sgaier, Grace Charles, Novi Quadrianto. Causal datasheet for datasets: An evaluation guide for real-world data analysis and data collection design using Bayesian Networks. Frontiers in Artificial Intelligence, p. 18. ISSN 2624-8212, 2021.
• Thomas Kehrenberg, Myles Bartlett, Oliver Thomas and Novi Quadrianto. Null-sampling for Interpretable and Fair Representations. European Conference on Computer Vision ECCV, Glasgow, UK, 2020.
• Viktoriia Sharmanska, Lisa Anne Hendricks, Trevor Darrell, Novi Quadrianto. Contrastive Examples for Addressing the Tyranny of the Majority. arXiv, 2020.