W3C home > Mailing lists > Public > public-web-adv@w3.org > December 2020

COWBIRD Proposal

From: Matthew Wilson <matt.wilson@nextroll.com>
Date: Tue, 8 Dec 2020 15:22:07 -0600
Message-ID: <CAFk2gV0y7MrU_8ToWtviXWcvZ7ubdRJkFacaVSMLZ=YAcyg+BA@mail.gmail.com>
To: public-web-adv@w3.org
Hello everyone,

I would like to share a proposal for how machine learning optimization
might be done in a privacy preserving way. The proposal is called COWBIRD
<https://github.com/AdRoll/privacy/blob/main/COWBIRD.md>, which stands for
Coordinated Optimization Without Big Resource Demands. It builds on SPURFOWL
<https://github.com/AdRoll/privacy/blob/main/SPURFOWL.md>, which we
discussed today, and is inspired by MURRE
<https://github.com/AdRoll/privacy/blob/main/MURRE.md>, which we discussed
a few weeks ago.

At a high level, the idea is for the browser to act as a federated learning
platform, allowing ad-tech companies to optimize toward customizable
objectives with customizable models and features.

Matt Wilson


Matthew Wilson

Staff Data Science Engineer

Received on Thursday, 10 December 2020 09:06:06 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:32:27 UTC