- Motivations and Context
- Scientific Objective
- the detection of an attack and the attackers [9],
- the suppression of the attacker's contribution in a model (unlearning).
- Strong skills in Machine Learning from a theoretical and practical point of view
- Mastering algorithms and python programing (scikit-learn, tensorflow, pytorch, etc.)
- Interest for cybersecurity
- Ability to read and understand scientific articles written in English
- Autonomy, rigor, curiosity, dedication, persistence and initiative are highly recommended qualities for completing a doctorate
- Research Master or engineering degree in computer science/applied mathematics with a specialization in Data-Science
- One or more research internships completed in the fields of Machine Learning and Deep Learning
-
Decentralised Federated Learning in a Dynamic
il y a 1 semaine
CEA - Commissariat à l'Energie Atomique Saclay, France**Domaine**: · Mathématiques, information scientifique, logiciel · **Contrat**: · Stage · **Intitulé de l'offre**: · Decentralised federated learning in a dynamic environment H/F · **Sujet de stage**: · Federated learning was introduced in 2016 by Google [1] as a new machine lear ...
-
Diversity & Inclusion Project Manager
il y a 1 semaine
AccorHotels Issy-les-Moulineaux, FranceCompany Description · Why work for Accor? · We are far more than a worldwide leader. We welcome you as you are and you can find a job and brand that matches your personality. We support you to grow and learn every day, making sure that work brings purpose to your life, so that du ...
-
Premium Midscale and Economy
il y a 1 semaine
AccorHotels Issy-les-Moulineaux, FranceCompany Description · Why work for Accor? · We are far more than a worldwide leader. We welcome you as you are and you can find a job and brand that matches your personality. We support you to grow and learn every day, making sure that work brings purpose to your life, so that du ...
-
Premium Midscale and Economy
il y a 6 jours
AccorHotels Issy-les-Moulineaux, FranceCompany Description · Why work for Accor? · We are far more than a worldwide leader. We welcome you as you are and you can find a job and brand that matches your personality. We support you to grow and learn every day, making sure that work brings purpose to your life, so that du ...
-
é Anglophone
il y a 1 semaine
FEM 21 Marseille e, France**Description entreprise**: · The Montessori 21 school in Marseille-Castellane opened its doors in September 2018. · It welcomes two atmospheres · - for 3-6 year olds, · - and one atmosphere for 6-11 year olds. · Member of the Montessori 21 network, bilingual French-English, the ...
-
Cemetery Operations Support Manager
il y a 1 jour
US American Battle Monuments Commission Paris, France**Duties**: · **Overseas Information** · Please note the following information related to this overseas assignment: · - The selectee will be required to obtain a diplomatic passport and appropriate visas (at the cost of the Agency) prior to assignment. · - The selectee will be re ...
-
Front-end Developer Advocate
il y a 3 jours
Jahia Paris, FranceWe are looking for a passionate Front-end developer Advocate to join our team in France. We have offices in Lyon / Paris. · Jahia started its life as a CMS platform, quite some time ago, and evolved since then into a digital experience platform (DXP) serving customers (such as Be ...
-
Team Lead Account Management
il y a 1 semaine
RTB House Paris, France**Who we are**: · Founded in 2012, RTB House offers marketing solutions that allow the biggest brands and media agencies to activate retargeting and branding campaigns. The company works with more than 3,500 advertisers (retail, travel, classified, telco, beauty, auto...). RTB Ho ...
-
Atlassian Expert
il y a 1 semaine
Modus Create Paris, France**Experience: Senior Level** · **Hello Are you ready to Work from Home and transform your career?** · We are looking for an** Atlassian Consultant Expert (Senior level)** with excellent consulting skills, who can consistently delight our customers and help grow our Atlassian prac ...
-
Team Lead Account Management
il y a 4 jours
RTB House Paris, France**Who we are**: · Founded in 2012, RTB House offers marketing solutions that allow the biggest brands and media agencies to activate retargeting and branding campaigns. The company works with more than 3,500 advertisers (retail, travel, classified, telco, beauty, auto...). RTB Ho ...
-
Support Engineer
il y a 1 semaine
Shift4 Paris, France**Overview**: · **Company Background** · The Giving Block, a Shift4 company, is the platform helping nonprofits fundraise more effectively from modern philanthropists. Pioneers of the "Crypto Philanthropy" movement, The Giving Block developed the leading solutions for cryptocurre ...
-
Global Team Icube
il y a 1 semaine
Imerys Paris, FranceThe Company · If you want to grow your skills and develop your career, find out more at · The Position · Global Team ICube (Continuous improvement) Programme Lead · Job Summary · Job Scope · Leading and facilitating Level 4, Wave 1, ICube Path Forward Implementation at different ...
-
Image Processing Expert
il y a 4 jours
PackageX Paris, France**About PackageX**: · PackageX is the world's first Headless Logistics Cloud. Hundreds of enterprises - across all industries - use PackageX's flexible first-to-last yard logistics APIs and infrastructure to digitize manual workflows, connect siloed point solutions and drive cust ...
-
Customer Success Manager
il y a 2 jours
Scaleway Paris, FranceDepuis 1999, Scaleway, filiale du groupe Iliad, figure parmi les leaders du cloud computing Européen. · Forte d'une équipe dynamique de près de 600 collaborateur(ice)s issus de tous horizons, notre entreprise offre un environnement de travail stimulant et international où se mêle ...
-
Customer Success Manager
il y a 2 jours
Scaleway Paris, FranceDepuis 1999, Scaleway, filiale du groupe Iliad, figure parmi les leaders du cloud computing Européen. · Forte d'une équipe dynamique de près de 600 collaborateur(ice)s issus de tous horizons, notre entreprise offre un environnement de travail stimulant et international où se mêle ...
-
Senior Full Stack Software Developer
il y a 3 jours
Jahia Paris, FranceJahia started its life as a CMS platform, quite some time ago, and evolved since then into a digital experience platform (DXP) serving customers (such as Ben &Jerry's, the European Parliament, Macom, Pôle Emploi) who either provide a public platform serving millions of unique vis ...
-
Talent Sourcing Intern
il y a 1 semaine
Pelico Paris, FranceAre you ready to be part of a groundbreaking company at the forefront of Industry 4.0? Pelico is on a mission to drive the transition of factories towards greater resilience, agility, and resource efficiency through data-driven collaboration. · Pelico, headquartered in Paris, Fra ...
-
Global Team Icube
il y a 1 semaine
Imerys Paris, FranceThe Company · If you want to grow your skills and develop your career, find out more at · The Position · Global Team ICube (Continuous improvement) Programme Member · Job Summary · Job Scope · Support facilitating Level 4, Wave 1, ICube Path Forward Implementation at different Fo ...
-
Staff Data Architect
il y a 1 semaine
Komodo Health Paris, France**We Breathe Life Into Data** · As we pursue these goals, it remains essential to us that we stay grounded in our values: be awesome, seek growth, deliver "wow," and enjoy the ride. At Komodo, you will be joining a team of ambitious, supportive Dragons with diverse backgrounds bu ...
-
Sustainability Intern
il y a 1 semaine
Brambles Group Paris, FranceCHEP helps move more goods to more people, in more places than any other organization on earth via our 300 million pallets, crates and containers. We employ 11,000 people and operate in more than 55 countries. Through our pioneering and sustainable share-and-reuse business model, ...
PhD "Machine Unlearning for attack mitigation in federated learning" M/F - Châtillon, France - Orange Business Services
Description
about the role
As a PhD student, you will work on a PhD thesis on the subject of: " Securing Federated Learning by Unlearning ".
Cybersecurity has become a major issue in an increasingly digital world. Indeed, cyberattacks are multiplying against both organizations and individuals. These attacks are increasingly sophisticated and now threaten Machine Learning (ML) models [1]. Researchers has thus shown the vulnerability of ML and the need to secure them [2].
Among these attacks against ML, we can cite the data poisoning attack where the attacker introduces or modifies the training data to corrupt the model of the victim. The attacker can thus introduce a backdoor in the model so that in the presence of a pattern the behavior of the model changes in accordance with his objective [3]. For example, for a self-driving car, the attacker could trick the AI by putting a green sticker on a "stop" sign so that it is recognized as a 70mph speed limit. To do this, he would simply have to modify the training dataset by introducing images of a "stop" sign with the label "stop" (legitimate data) accompanied by these same images but after adding a small green square with the label "70" (malicious data).
While it is obviously difficult for an attacker to gain privileged access to a training dataset, it is on the other hand easier for him to inject his malicious data in a collaborative learning setting.
In the field of collaborative learning, federated learning (FL) is a new paradigm of ML allowing participants to collaborate in the training of a global model while preserving the data privacy [4]. Thus, instead of sharing their data, participants share their model parameters. This approach arouses a large interest, especially at Orange. However, it is more vulnerable to the risk of poisoning attacks due to the probability of having an attacker among the diversity of participants [5].
One of the particularities of backdoors is that they are introduced by a subtle modification of the training set. It is therefore difficult to detect them during the training phase and they are often discovered during the production phase. Training being an expensive process, deleting the model and restarting from scratch is not an option. It is therefore necessary to unlearn the contribution of the malicious participants [6,7,8].
The objective of this thesis is therefore to study and propose new unlearning algorithms in the federated learning setting. For this, it will be necessary to deeply understand the attack mechanisms against FL and ML to develop a framework to reproduce these attacks. The focus will be on the backdoor attacks. The main challenges to overcome are:
The results of this work can then be used to secure Orange's federated learning platform "OrangeFL".
about you
Expected Skills (scientific, technical and personal)
Educational background and diploma
Desired experiences
additional information
Cybersecurity as well as artificial intelligence is experiencing an unprecedented global interest. These two fields are not only fascinating but also full of opportunities. The doctoral student will be able to explore a new and growing field of research where the security of AI models is at the heart of the problem.
She/he will also have access to data from connected vehicles (V2X) to test her/his approaches. In addition, during the thesis, the candidate will benefit from a stimulating environment for research.
References
department
Orange Innovation brings together the research and innovation activities and expertise of the Group's entities and countries. We work every day to ensure that Orange is recognized as an innovative operator by its customers, and we create value for the Group and the Brand in each of our projects. With 720 researchers, thousands of marketers, developers, designers and data analysts, it is the expertise of our 6,000 employees that fuels this ambition every day.
Orange Innovation anticipates technological breakthroughs and supports the Group's countries and entities in making the best technological choices to meet the needs of our consumer and business customers.
Within Orange Innovation, you will join the MORE (Mathematical Models for Optimization and peRformance Evaluation) team which has around ten permanent engineers/researchers whose mission is to develop models and techniques to optimize the quality and performance of Orange group services. The team is also made up of around ten doctoral students, apprentices, and interns. You will thus be part of a research ecosystem alongside operational units, with the aim of developing algorithms at the cutting edge of innovation.
contract
Thesis