Project Name
FELToken
Project Category
Build & Integrate
Proposal Earmark
General
Proposal Description
The final product of FELToken is a platform for running machine learning across distributed data sets without compromising the privacy of the data. Previously we created our tool for running data provider clients and web applications for managing the projects. However, we decided to change the architecture and focus on entirely using Ocean protocol. We want to use Ocean’s compute-to-data for all computation. FELToken will then act orchestration tool on top of the Ocean protocol. It will provide necessary algorithms so that users can easily select datasets and train machine learning models on top of them.
During development, we focus mainly on making the process of setting up the federated learning project as simple as possible. So that data scientists and interested parties can train their models without any extra knowledge about smart contracts and blockchain. Once the core components are working, there are many possible ways for further extension.
Grant Deliverables
- Working demo of new architecture (full usage of Ocean protocol)
- Training local models on selected Ocean datasets
- Algorithm for aggregating local models into global model
- Web application for starting the algorithms and monitoring the progress
- Documentation for new architecture
- Blog post about our progress
Project Description
FELToken is connecting owners of the data with scientists to train their machine learning models while preserving the privacy of the data. Current federated learning solutions rely on a central server. We are building a product with the simplest user experience possible while being decentralized. This allows anonymous participation of data providers and prevents malicious activities. Data providers get rewards for sharing their data and resulting models can be further sold. Exchange of all models is encrypted so only authorized parties can use the model. We would like to further improve our tool by connecting with Ocean infrastructure.
Final Product
Platform for a decentralized and more secure solution for federated learning while anonymizing data providers. Allowing data scientists to do machine learning on decentralized data without compromising its privacy.
Value Add Criteria
With our previous architecture, we previously finished MVP: https://youtu.be/uoBl2yeO7hY and integration with Ocean protocol: https://twitter.com/FELToken/status/1508499624918589447.
We decided to change the old architecture and fully focus on Ocean protocol. With the new architecture, all users of FELToken will also be users of Ocean protocol. FELToken will allow users to efficiently perform train machine learning models across multiple datasets published on Ocean. The ability to easily train machine learning models will drive more users to Ocean protocol. It will increase the value of data already stored on Ocean protocol. Moreover, with the usage of compute-to-data, we can also preserve the privacy of the data.
Core Team
Břetislav Hájek
Role: developer
Relevant Credentials:
- GitHub: Breta01 (Břetislav Hájek) · GitHub
- LinkedIn: https://www.linkedin.com/in/břetislav-hájek-75167111b
Background/Experience:
- AI developer: https://inventives.ai/
- PhD student at: https://www.nus.edu.sg/
Filip Masár
Role: developer
Relevant Credentials:
- GitHub: FilipMasar (Filip Masar) · GitHub
- LinkedIn: https://www.linkedin.com/in/filip-masar-776a0a174/
Background/Experience:
- Kernel KB4 fellow https://kernel.community/
- Co-founder at http://avtr.gg/
- Front-end developer at https://avast.com/
Martin Ondejka
Role: developer
Relevant Credentials:
Background/Experience:
- Python developer at https://www.kiwi.com/us/
- Software Engineer at https://vacuumlabs.com/
Funding Requested
10000
Minimum Funding Requested
Wallet Address
0x77edDB82E5e9901aA494825bC362fA93120B892c