Abstract In the high-stakes realm of critical care, where daily decisions are crucial and clear communication is paramount, comprehending the rationale behind Artificial Intelligence (AI)-driven decisions appears essential. While AI has the potential to improve decision-making, its complexity can hinder comprehension and adherence to its recommenda...
This paper explores the theoretical foundations of fair regression under the constraint of demographic parity within the unawareness framework, where disparate treatment is prohibited, extending existing results where such treatment is permitted. Specifically, we aim to characterize the optimal fair regression function when minimizing the quadratic...
Published in Journal of neurodevelopmental disorders
Sharing and fairness are important prosocial behaviors that help us navigate the social world. However, little is known about how and whether individuals with Williams Syndrome (WS) engage in these behaviors. The unique phenotype of individuals with WS, consisting of high social motivation and limited social cognition, can also offer insight into t...
Machine learning techniques are increasingly used for high-stakes decision-making, such as college admissions, loan attribution or recidivism prediction. Thus, it is crucial to ensure that the models learnt can be audited or understood by human users, do not create or reproduce discrimination or bias, and do not leak sensitive information regarding...
We address the problem of performing regression while ensuring demographic parity, even without access to sensitive attributes during inference. We present a general-purpose post-processing algorithm that, using accurate estimates of the regression function and a sensitive attribute predictor, generates predictions that meet the demographic parity ...
Des études récentes ont introduit des tech- niques de compression efficaces pour les grands modèles de langage (LLMs) via la quantifica- tion post-entraînement ou la représentation des poids en bits faibles. Bien que les poids quan- tifiés offrent une efficacité de stockage et per- mettent une inférence plus rapide, les travaux existants ont indiqu...
While machine learning (ML) has been extensively used in Massive Open Online Courses (MOOCs) to predict whether learners are at risk of dropping-out or failing, very few work has investigated the bias or possible unfairness of the predictions generated by these models. This is however important, because MOOCs typically engage very diverse audiences...
International audience
Recent studies introduced effective compres- sion techniques for Large Language Models (LLMs) via post-training quantization or low- bit weight representation. Although quantized weights offer storage efficiency and allow for faster inference, existing works have indicated that quantization might compromise perfor- mance and exacerbate biases in LL...
This thesis explores the importance of strengthening compliance with regulatory frameworks like the European General Data Protection Regulation (GDPR) in relation to data privacy and fairness in the field of voice biometrics focusing specifically on Automatic Speaker Verification (ASV). Through the use of cryptographic techniques, data perturbation...
