Ruta Binkyte

ruta.binkyte@gmail.com

I am a doctoral Machine Learning Ethics researcher at Inria Saclay-Ile-de-France, LIX École Polytechnique and IP Paris and an Associate Professor at Aivancity School for Technolofy, Business and Society. I consider my diverse background in cultural anthropology, history, and data science as my main strength and leverage it to address the complex interdisciplinary questions in ethical ML.

My Ph.D. work at the Institute Polytechnique de Paris, funded by the ERC Hypatia grant, is at the intersection of fairness, privacy, and causality and investigates practical and theoretical tools for developing ethical machine learning implementations.

My research is published or forthcoming in peer-reviewed journals, including Journal of Information Processing and Management and NeurIPS Workshop on Algorithmic Fairness Through The Lens of Causality and Privacy .

I have co-designed (with Sami Zhioua ) and teach the course on ML Fairness at Aivancity, an interdisciplinary AI and business school in Paris. I am an author of learning material on Algorithmic Media (a project in partnership with NGO Meno Avilys and British Council Lithuania) for high school children and a Data Science coach (in collaboration with Teachers Lead Tech) for primary schools in Lithuania.

I have contributed to various research projects and initiatives, including being a member of the Scientific Programme Committee at The 10th IEEE Swiss Conference on Data Science (SDS2023) , co-organizing (with Bilel Benbouzid ) the Interdisciplinary Workshop on Fairness in AI , bringing together more than 50 researchers from social science, computer schience and law, co-rganizing The first workshop at COMETE Inria on Ethical AI.

I am also proud to hold a Google WTM Scholar Award for 2020-2021.

news

Jan 27, 2023	UPCOMING: I am going to give a talk based on our paper ‘Gender and sex bias in COVID-19 epidemiological data through the lenses of causality’ at a WiMLDS (Women in Machine Learning and Data Science) Paris event in March 2023 (details to be announced).
Oct 22, 2022	I am giving a talk on Causality for Fairness at The 7th Franco-Japanese Cybersecurity Workshop at Keio University Tokyo. FranceJapanICST
Sep 1, 2022	I am proud to join the faculty at Aivancity School for Business and Technology as an Associate Professor . I will be teaching a Fair ML course for Master’s students there.

selected publications

Causality

Gender and sex bias in COVID-19 epidemiological data through the lenses of causality

Natalia Dı́az-Rodrı́guez, Rūta Binkytė, Wafae Bakkali, and 5 more authors

Information Processing & Management 2023
Causality

Causal Discovery for Fairness

Rūta Binkytė, Karima Makhlouf, Carlos Pinzón, and 2 more authors

arXiv preprint arXiv:2206.06685 2022

Abs HTML PDF

It is crucial to consider the social and ethical consequences of AI and ML based decisions for the safe and acceptable use of these emerging technologies. Fairness, in particular, guarantees that the ML decisions do not result in discrimination against individuals or minorities. Identifying and measuring reliably fairness/discrimination is better achieved using causality which considers the causal relation, beyond mere association, between the sensitive attribute (e.g. gender, race, religion, etc.) and the decision (e.g. job hiring, loan granting, etc.). The big impediment to the use of causality to address fairness, however, is the unavailability of the causal model (typically represented as a causal graph). Existing causal approaches to fairness in the literature do not address this problem and assume that the causal model is available. In this paper, we do not make such assumption and we review the major algorithms to discover causal relations from observable data. This study focuses on causal discovery and its impact on fairness. In particular, we show how different causal discovery approaches may result in different causal models and, most importantly, how even slight differences between causal models can have significant impact on fairness/discrimination conclusions. These results are consolidated by empirical analysis using synthetic and standard fairness benchmark datasets. The main goal of this study is to highlight the importance of the causal discovery step to appropriately address fairness using causality.