Randomized Response (RR) is a protocol designed to collect and analyze categorical data with local differential privacy guarantees. It has been used as a building block of mechanisms deployed by Big tech companies to collect app or web users' data. Each user reports an automatic random alteration of their true value to the analytics server, which then estimates the histogram of the true unseen values of all users using a debiasing rule to compensate for the added randomness. A known issue is that the standard debiasing rule can yield a vector with negative values (which can not be interpreted as a histogram), and there is no consensus on the best fix. An elegant but slow solution is the Iterative Bayesian Update algorithm (IBU), which converges to the Maximum Likelihood Estimate (MLE) as the number of iterations goes to infinity. This paper bypasses IBU by providing a simple formula for the exact MLE of RR and compares it with other estimation methods experimentally to help practitioners decide which one to use.

International audience

One of the main concerns about fairness in machine learning (ML) is that, in order to achieve it, one may have to trade off some accuracy. To overcome this issue, Hardt et al. proposed the notion of equality of opportunity (EO), which is compatible with maximal accuracy when the target label is deterministic with respect to the input features. In the probabilistic case, however, the issue is more complicated: It has been shown that under differential privacy constraints, there are data sources for which EO can only be achieved at the total detriment of accuracy, in the sense that a classifier that satisfies EO cannot be more accurate than a trivial (random guessing) classifier. In our paper we strengthen this result by removing the privacy constraint. Namely, we show that for certain data sources, the most accurate classifier that satisfies EO is a trivial classifier. Furthermore, we study the trade-off between accuracy and EO loss (opportunity difference), and provide a sufficient condition on the data source under which EO and non-trivial accuracy are compatible.

A blockchain is a distributed hierarchical data structure. Widely-used applications of blockchain include digital currencies such as Bitcoin and Ethereum. This paper proposes an algorithmic approach to analyze the efficiency of a blockchain as a function of the number of blocks and the average synchronization delay. The proposed algorithms consider a random network model that characterizes the growth of a tree of blocks by adhering to a standard protocol. The model is parametric on two probability distribution functions governing block production and communication delay. Both distributions determine the synchronization efficiency of the distributed copies of the blockchain among the so- called workers and, therefore, are key for capturing the overall stochastic growth. Moreover, the algorithms consider scenarios with a fixed or an unbounded number of workers in the network. The main result illustrates how the algorithms can be used to evaluate different types of blockchain designs, e.g., systems in which the average time of block production can match the average time of message broadcasting required for synchronization. In particular, this algorithmic approach provides insight into efficiency criteria for identifying conditions under which increasing block production has a negative impact on the stability of a blockchain. The model and algorithms are agnostic of the blockchain’s final use, and they serve as a formal framework for specifying and analyzing a variety of non-functional properties of current and future blockchains.