Analysis of client profitability in a distributed federated learning system using Blockchain H / F – Stage – Palaiseau

Within CEA Tech, CEA’s technology research department, the List Institute (CEA List) is dedicated to intelligent digital systems. We have the experience that results from a culture of innovation and our mission is to produce and transfer useful technologies to our industrial partners.

LICIA (Laboratoire systèmes d’Information de Confiance, Intelligents, Auto-organisants) is CEA List’s laboratory dedicated to the study of “Trusted Distributed Systems”, and in particular, blockchain technology.

Analysis of client profitability in a distributed federated learning system using Blockchain H / F

The goal of this internship is to analyze user profitability in a federated learning setup.

Machine learning is today one of the main fields of study in theoretical computing. It has many different applications and is useful in our daily life, for example, in search engines (Google), in translations (Deepl), for feed recommendations (Facebook, Youtube). Some medical services also use machine learning for the analysis of medical data, the control of inappropriate data, the estimation of disease progress, the effective follow-up of patients, etc. These services typically collect all the data and use their training algorithms to produce very specific results.

However, in part, due to some privacy reasons, it is increasingly interesting to consider federated learning (FL), for example, see Yang et al. (2018), Li et al. (2020). In FL, data is always on the users side. A server is responsible for sending users a learning model; users update the model using their data and learning algorithms, then send the update to the server. The server then adds them to a new model, which is sent back to users, and the process continues until the end of the learning.

CEA List’s Carnot Fanstastyc project aims to fully decentralize FL’s server role and thus build distributed federated learning applications. For this, blockchain technology is used to guarantee transparency, order and agreement on the sequence of updates and models.

In such systems, the role of all users is essential. If some users decide not to participate, the system may fail. Therefore, it is important to ensure that different users have the incentive to participate in the system.

For this, the selected candidate will analyze the profitability of the different actors, especially the profitability of the clients.
In more detail, the objective is (i) to show the existence of the different (Nash) equilibria of the game that represents the interactions of the users, (ii) to calculate these equilibria if they exist, and (iii) if possible, to propose mechanisms of incentive that induce the existence of a unique balance in which the best interest of users is consistent with their participation in federated learning processes.
The internship is located on the Paris-Saclay campus, at LICIA, CEA List’s laboratory dedicated to distributed systems and blockchain technologies.

The candidate must be in “Master 2” or equivalent, in computer science or related fields.
The candidate is expected to know theoretical computing and especially (algorithmic) game theory. Knowledge of federated learning and / or blockchain technologies is appreciated, but not mandatory.

The internship is open to French and English speakers.

Rank Math Seo Pro Weadown, Wordfence Premium Nulled, Yoast Nulled, PHP Script, Fs Poster Plugin Nulled, Astra Pro Nulled,Woodmart Theme Nulled, Wpml Nulled, Avada 7.4 Nulled, Woodmart Theme Nulled, PW WooCommerce Gift Cards Pro Nulled, Elementor Pro Weadown, Newspaper – News & WooCommerce WordPress Theme, Nulledfire, Slider Revolution Nulled, Elementor Pro Weadown, Jnews 8.1.0 Nulled, WeaPlay, Business Consulting Nulled, WP Reset Pro, Newspaper 11.2, Flatsome Nulled, Woocommerce Custom Product Ad, Premium Addons for Elementor, Jannah Nulled, Consulting 6.1.4 Nulled, Plugins, WordPress Theme, Dokan Pro Nulled

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker.