Image 1 of 1: ‘Two tables with data are moved to a central location’
Figure 2
Image 1 of 1: ‘Names are censored before the datasets are sent to a central place’
Figure 3
Image 1 of 1: ‘In federated analysis, an aggregation of the local data is sent to a central point. At the central point the local aggregations are combined. This can also be a small step in a larger analysis.’
Figure 4
Image 1 of 1: ‘In federated learning only the gradients of models are shared’
Figure 5
Image 1 of 1: ‘An example of gradient leakage. The order might not be correct but the images are still very close to the originals.’
Figure 6
Image 1 of 1: ‘In secure multiparty computation parties collaboratively perform an analysis while holding only encrypted pieces of the data’
Figure 7
Image 1 of 1: ‘Mees, Sara and Noor distibute their secret shares’
Figure 8
Image 1 of 1: ‘Mees, Sara and Noor sum their secret shares’
Figure 9
Image 1 of 1: ‘Mees, Sara and Noor add their shares together for the final result’
Figure 10
Image 1 of 1: ‘Differential privacy sometimes replaces a subset of the data with random values’
Figure 11
Image 1 of 1: ‘Horizontal and vertical partitioning refers to how data is separated’
Image 1 of 1: ‘High level overview of the vantage6 infrastructure. Client(s) and node(s) communicate through the Server. Nodes are able to communicate directly with each other when the optional VPN feature is enabled.’
Figure 7
Figure 8
Image 1 of 1: ‘Federated analysis orchestrated by vantage6’
Figure 9
Image 1 of 1: ‘vantage6 central and federated tasks.’