Existing Work by Other PhDs

The following algorithms & tools have already been implemented by other PhDs. Please make use of these.

Federated Learning algorithms

Privacy preserving n-party Scalar product protocol

• Language: java
• Use-case: independently or within vantage6
• Developer: Florian
• Repo: https://github.com/MaastrichtU-CDS/n-scalar-product-protocol
• Scientific article: https://ieeexplore.ieee.org/document/10024342

An algorithm to calculate the scalar product for n-parties.

VertiBayes

• Language: java, python (vantage6 integration)
• Use-case: independently or within vantage6
• Developer: Florian
• Repo: https://github.com/MaastrichtU-CDS/vertibayes
• Preprint: https://arxiv.org/abs/2210.17228

An algorithm for training bayesian networks in a federated setting. Can handle vertical, horizontal and hybrid splits of the data. Can handle an arbitrary amount of parties. Can handle missing data.

FBNE

• Language: java, python (vantage6 integration)
• Use-case: independently or within vantage6
• Developer: Florian
• Repo: https://github.com/MaastrichtU-CDS/bayesianEnsemble
• Article: https://ieeexplore.ieee.org/document/10306230
• Free version: https://arxiv.org/abs/2402.12142

Federated Learning infrastructure

ARGOS infra

• Language: shell
• Use-case: federated deep learning
• Developer: Ananya et al.
• Repo: https://github.com/MaastrichtU-CDS/argos-infrastructure

Adaptation of the vantage6 federated learning infrastructure with an added computing node.

FAIRification tools

FAIRnotator

• Language: mostly python
• Use-case: triplifier method
• Developer: Johan, Shashank et al.
• Repo: https://github.com/MaastrichtU-CDS/EPND-FAIRification

Semi-automatic tool for data harmonisation and metadata creation.

FAIRifier

• Language: python
• Use-case: R2RML method
• Developer: Jasper, Aiara et al.
• Repo: https://github.com/MaastrichtU-CDS/fairifier2

Semi-automatic tool for data harmonisation.