Ulrich Aivodji

Eliott Baltz

Maîtrise recherche - École de technologie suprérieure

eliott.baltz@mila.quebec

maryam.babaie@gmail.com Babaei

Doctorat - École de technologie suprérieure

maryam.babaei@mila.quebec

meghana.bhange@mila.quebec

Meghana Bhange

Maîtrise recherche - École de technologie suprérieure

Co-superviseur⋅e :

Samira Ebrahimi Kahou

Doctorat - École de technologie suprérieure

Co-superviseur⋅e :

Samira Ebrahimi Kahou

patrik.kenfack@mila.quebec

Google Scholar

Publications

Probabilistic Dataset Reconstruction from Interpretable Models

Julien Ferry

Sébastien Gambs

Marie-José Huguet

Mohamed Siala

Interpretability is often pointed out as a key requirement for trustworthy machine learning. However, learning and releasing models that are… (voir plus) inherently interpretable leaks information regarding the underlying training data. As such disclosure may directly conflict with privacy, a precise quantification of the privacy impact of such breach is a fundamental problem. For instance, previous work have shown that the structure of a decision tree can be leveraged to build a probabilistic reconstruction of its training dataset, with the uncertainty of the reconstruction being a relevant metric for the information leak. In this paper, we propose of a novel framework generalizing these probabilistic reconstructions in the sense that it can handle other forms of interpretable models and more generic types of knowledge. In addition, we demonstrate that under realistic assumptions regarding the interpretable models' structure, the uncertainty of the reconstruction can be computed efficiently. Finally, we illustrate the applicability of our approach on both decision trees and rule lists, by comparing the theoretical information leak associated to either exact or heuristic learning algorithms. Our results suggest that optimal interpretable models are often more compact and leak less information regarding their training data than greedily-built ones, for a given accuracy level.

2024-01-01

SaTML (publié)

openreview.net

Fairness Under Demographic Scarce Regime

Patrik Joslin Kenfack

Samira Ebrahimi Kahou

Most existing works on fairness assume the model has full access to demographic information. However, there exist scenarios where demographi… (voir plus)c information is partially available because a record was not maintained throughout data collection or due to privacy reasons. This setting is known as demographic scarce regime. Prior research have shown that training an attribute classifier to replace the missing sensitive attributes (proxy) can still improve fairness. However, the use of proxy-sensitive attributes worsens fairness-accuracy trade-offs compared to true sensitive attributes. To address this limitation, we propose a framework to build attribute classifiers that achieve better fairness-accuracy trade-offs. Our method introduces uncertainty awareness in the attribute classifier and enforces fairness on samples with demographic information inferred with the lowest uncertainty. We show empirically that enforcing fairness constraints on samples with uncertain sensitive attributes is detrimental to fairness and accuracy. Our experiments on two datasets showed that the proposed framework yields models with significantly better fairness-accuracy trade-offs compared to classic attribute classifiers. Surprisingly, our framework outperforms models trained with constraints on the true sensitive attributes.

2023-07-24

ArXiv (preprint)

openreview.net

Learning Hybrid Interpretable Models: Theory, Taxonomy, and Methods

Julien Ferry

gabriel laberge

A hybrid model involves the cooperation of an interpretable model and a complex black box. At inference, any input of the hybrid model is as… (voir plus)signed to either its interpretable or complex component based on a gating mechanism. The advantages of such models over classical ones are two-fold: 1) They grant users precise control over the level of transparency of the system and 2) They can potentially perform better than a standalone black box since redirecting some of the inputs to an interpretable model implicitly acts as regularization. Still, despite their high potential, hybrid models remain under-studied in the interpretability/explainability literature. In this paper, we remedy this fact by presenting a thorough investigation of such models from three perspectives: Theory, Taxonomy, and Methods. First, we explore the theory behind the generalization of hybrid models from the Probably-Approximately-Correct (PAC) perspective. A consequence of our PAC guarantee is the existence of a sweet spot for the optimal transparency of the system. When such a sweet spot is attained, a hybrid model can potentially perform better than a standalone black box. Secondly, we provide a general taxonomy for the different ways of training hybrid models: the Post-Black-Box and Pre-Black-Box paradigms. These approaches differ in the order in which the interpretable and complex components are trained. We show where the state-of-the-art hybrid models Hybrid-Rule-Set and Companion-Rule-List fall in this taxonomy. Thirdly, we implement the two paradigms in a single method: HybridCORELS, which extends the CORELS algorithm to hybrid modeling. By leveraging CORELS, HybridCORELS provides a certificate of optimality of its interpretable component and precise control over transparency. We finally show empirically that HybridCORELS is competitive with existing hybrid models, and performs just as well as a standalone black box (or even better) while being partly transparent.

2023-03-08

ArXiv (prépublication)

arxiv.org

Fooling SHAP with Stealthily Biased Sampling

gabriel laberge

Satoshi Hara

Mario Marchand

Foutse Khomh

SHAP explanations aim at identifying which features contribute the most to the difference in model prediction at a speciﬁc input versus a … (voir plus)background distribution. Recent studies have shown that they can be manipulated by malicious adversaries to produce arbitrary desired explanations. However, existing attacks focus solely on altering the black-box model itself. In this paper, we propose a complementary family of attacks that leave the model intact and manipulate SHAP explanations using stealthily biased sampling of the data points used to approximate expectations w.r.t the background distribution. In the context of fairness audit, we show that our attack can reduce the importance of a sensitive feature when explaining the difference in outcomes between groups, while remaining undetected. These results highlight the manipulability of SHAP explanations and encourage auditors to treat post-hoc explanations with skepticism.

2023-02-01

ICLR.cc/2023/Conference (poster)

openreview.net

Leveraging Integer Linear Programming to Learn Optimal Fair Rule Lists

Julien Ferry

Sébastien Gambs

Marie-José Huguet

Mohamed

Siala

2022-06-10

Integration of Constraint Programming, Artificial Intelligence, and Operations Research (publié)

Washing The Unwashable : On The (Im)possibility of Fairwashing Detection

A. Shamsabadi

Mohammad Yaghini

Natalie Dullerud

Sierra Calanda Wyllie

Aisha Alaagib

Sébastien Gambs

Nicolas Papernot

Local Data Debiasing for Fairness Based on Generative Adversarial Training

François Bidet

Sébastien Gambs

Rosin Claude Ngueveu

Alain Tapp

The widespread use of automated decision processes in many areas of our society raises serious ethical issues with respect to the fairness o… (voir plus)f the process and the possible resulting discrimination. To solve this issue, we propose a novel adversarial training approach called GANSan for learning a sanitizer whose objective is to prevent the possibility of any discrimination (i.e., direct and indirect) based on a sensitive attribute by removing the attribute itself as well as the existing correlations with the remaining attributes. Our method GANSan is partially inspired by the powerful framework of generative adversarial networks (in particular Cycle-GANs), which offers a flexible way to learn a distribution empirically or to translate between two different distributions. In contrast to prior work, one of the strengths of our approach is that the sanitization is performed in the same space as the original data by only modifying the other attributes as little as possible, thus preserving the interpretability of the sanitized data. Consequently, once the sanitizer is trained, it can be applied to new data locally by an individual on their profile before releasing it. Finally, experiments on real datasets demonstrate the effectiveness of the approach as well as the achievable trade-off between fairness and utility.

2021-03-14

Algorithms (publié)

arxiv.org

Fairwashing: the risk of rationalization