Internet of Medical Things (IoMTs) is revolutionizing the healthcare industry regarding how diagnosis process takes place, how treatment is provided, and how public health policies are made. A real-world use case of IoMTs is to investigate how infectious diseases, e.g. COVID-19, spread in a population through social events. In this use case, people’s social contact records in certain venues are collected by sensors and saved locally; pandemic modellers, as third-party vendors, are desired to construct social contact network based on contacts records, and to simulate the process of disease transmission over the contact network by transmission modelling; results from the simulation will be provided to authorities for policymaking and pandemic control. However, concerns are raised on data breaches from modellers. In reality, sharing the data in clear with modellers is not allowed by regulations for the sake of privacy. In this work, we will be addressing the contradiction between data privacy and usability when vendors are involved in IoMTs. We propose a secure cloud-edge computing architecture based on an efficient fully homomorphic encryption (FHE) scheme. This architecture allows vendors to securely and “blindly” process medical data without compromising the quality of their service. Moreover, we apply the proposed architecture to the use case of pandemic modelling. By comparisons with a differential privacy-based solution, we demonstrate the favorable feasibility, accuracy and security of the proposed solution.

AI-as-a-Service has emerged as an important trend for supporting the growth of the digital economy. Digital service providers make use of their vast amount of customer data to train AI models (such as image recognition, financial modelling and pandemic modelling etc) and offer them as a service on the cloud. While there are convincing advantages for using such third-party models, the fact that model users are required to upload their data to the cloud is bound to raise serious privacy concerns, especially in the face of increasingly stringent privacy regulations and legislation. To promote the adoption of AI-as-a-Service while addressing privacy issues, we propose a practical approach for constructing privacy-enhanced neural networks by designing an efficient implementation of fully homomorphic encryption. With this approach, an existing neural network can be converted to process FHE-encrypted data and produce encrypted output which are only accessible by the model users, and more importantly, within an operationally acceptable time (e.g. within 1 second for facial recognition in typical border control systems). Experimental results show that in many practical tasks such as facial recognition, text classification and so on, we obtained the state-of-the-art inference accuracy in less than one second on a 16 cores CPU.

This paper introduces the targeted sampling model in optimal auction design. In this model, the seller may specify a quantile interval and sample from a buyer's prior restricted to the interval. This can be interpreted as allowing the seller to, for example, examine the top 40% bids from previous buyers with the same characteristics. The targeting power is quantified with a parameter Δ ∈ [0, 1] which lower bounds how small the quantile intervals could be. When Δ = 1, it degenerates to Cole and Roughgarden's model of i.i.d. samples; when it is the idealized case of Δ = 0, it degenerates to the model studied by [7]. For instance, for n buyers with bounded values in [0, 1], ~O(ε-1) targeted samples suffice while it is known that at least ~Ømega(n ε-2) i.i.d. samples are needed. In other words, targeted sampling with sufficient targeting power allows us to remove the linear dependence in n, and to improve the quadratic dependence in ε-1 to linear. In this work, we introduce new technical ingredients and show that the number of targeted samples sufficient for learning an ε-optimal auction is substantially smaller than the sample complexity of i.i.d. samples for the full spectrum of Δ ∈ [0, 1). Even with only mild targeting power, i.e., whenever Δ = o(1), our targeted sample complexity upper bounds are strictly smaller than the optimal sample complexity of i.i.d. samples.

We consider a generalization of the third degree price discrimination problem studied in [4](Bergemann et al., 2015), where an intermediary between the buyer and the seller can design market segments to maximize any linear combination of consumer surplus and seller revenue. Unlike in [4], we assume that the intermediary only has partial information about the buyer's value. We consider three different models of information, with increasing order of difficulty. In the first model, we assume that the intermediary's information allows him to construct a probability distribution of the buyer's value. Next we consider the sample complexity model, where we assume that the intermediary only sees samples from this distribution. Finally, we consider a bandit online learning model, where the intermediary can only observe past purchasing decisions of the buyer, rather than her exact value. For each of these models, we present algorithms to compute optimal or near optimal market segmentation.

We consider the problem of learning optimal reserve price in repeated auctions against non-myopic bidders, who may bid strategically in order to gain in future rounds even if the single-round auctions are truthful. Previous algorithms, e.g., empirical pricing, do not provide non-trivial regret rounds in this setting in general. We introduce algorithms that obtain small regret against non-myopic bidders either when the market is large, i.e., no bidder appears in a constant fraction of the rounds, or when the bidders are impatient, i.e., they discount future utility by some factor mildly bounded away from one. Our approach carefully controls what information is revealed to each bidder, and builds on techniques from differentially private online learning as well as the recent line of works on jointly differentially private algorithms.