… learned indexes—a trending technique that repurposes machine learning models as indexing
… We consider the problem of private training and inference of learned indexes on static data…

… attackers aim to learn a non-colluding party’s private information … differentially private marginal
distributions of crossdataset LA pairs in Section 5.4.1. Then, we present a novel tree index …

… Several solutions, based on Federated Learning transfer and update models by … In this
paper, we propose AdaDpFed, an adaptive federated differential private protocol in the non-IID …

… For the sake of clarity, we will drop specific indexing of communication rounds for now. … , the
differentially private model almost reaches accuracies of the non-differential private one. This …

… In this paper, we focus on spatial data, ie, any multi-dimensional data that can be indexed
… definition: it ensures that what can be learned from the released data does not substantially …

… To emphasize, we first define the desired fairness criteria for differentially private learning. …
This section reports our evaluation of the fair differentially private learning on benchmark …

… In this section, we show that Algorithm 2 is εp-differentially private. This proof requires stronger
assumptions on the loss function than were required in Theorem 6. In certain cases, some …

… of learned indexes, we propose a lossy parameter compression strategy for data-based
partitioning learned indexes … Differentially private highdimensional data publication via grouping …

… privacy analysis for differentially private deep learning, as elaborated later in this section. …
Each review has been encoded as a list of word indexes. Words are indexed by their overall …

… By the comparison of GHS Ad (adjusting EU of G-HS point by point to make its ASR value
close to that of G-GS) to G-GS, we find out that in the average sense the indexes, WTD, HOP, …