The latest pandemic COVID-19 brought governments worldwide to use various containment measures to control its spread, such as contact tracing, social distance regulations, and curfews. Epidemiological simulations are commonly used to assess the impact of those policies before they are implemented. Unfortunately, the scarcity of relevant empirical data, specifically detailed social contact graphs, hampered their predictive accuracy. As this data is inherently privacy-critical, a method is urgently needed to perform powerful epidemiological simulations on real-world contact graphs without disclosing any sensitive information. In this work, we present RIPPLE, a privacy-preserving epidemiological modeling framework enabling standard models for infectious disease on a population’s real contact graph while keeping all contact information locally on the participants’ devices. As a building block of independent interest, we present PIR-SUM, a novel extension to private information retrieval for secure download of element sums from a database. Our protocols are supported by a proof-of-concept implementation, demonstrating a 2-week simulation over half a million participants completed in 7 minutes, with each participant communicating less than 50 KB.

While secure multi-party computation (MPC) protects the privacy of inputs and intermediate values of a computation, differential privacy (DP) ensures that the output itself does not reveal too much about individual inputs. For this purpose, MPC can be used to generate noise and add this noise to the output. However, securely generating and adding this noise is a challenge considering real-world implementations on finite-precision computers, since many DP mechanisms guarantee privacy only when noise is sampled from continuous distributions requiring infinite precision.

The advent of transformers has brought about significant advancements in traditional machine learning tasks. However, their pervasive deployment has raised concerns about the potential leakage of sensitive information during inference. Existing approaches using secure multiparty computation (MPC) face limitations when applied to transformers due to the extensive model size and resource-intensive matrix-matrix multiplications. In this paper, we present BOLT, a privacy-preserving inference framework for transformer models that supports efficient matrix multiplications and nonlinear computations. Combined with our novel machine learning optimizations, BOLT reduces the communication cost by 10.91×. Our evaluation on diverse datasets demonstrates that BOLT maintains comparable accuracy to floating-point models and achieves 4.8-9.5× faster inference across various network settings compared to the state-of-the-art system.

Over the last two years, governments all over the world have used a variety of containment measures to control the spread of \covid, such as contact tracing, social distance regulations, and curfews. Epidemiological simulations are commonly used to assess the impact of those policies before they are implemented in actuality. Unfortunately, their predictive accuracy is hampered by the scarcity of relevant empirical data, concretely detailed social contact graphs. As this data is inherently privacy-critical, there is an urgent need for a method to perform powerful epidemiological simulations on real-world contact graphs without disclosing sensitive information.

Clustering is an unsupervised machine learning technique that outputs clusters containing similar data items. In this work, we investigate privacy-preserving density-based clustering which is, for example, used in financial analytics and medical diagnosis. When (multiple) data owners collaborate or outsource the computation, privacy concerns arise. To address this problem, we design, implement, and evaluate the first practical and fully private density-based clustering scheme based on secure two-party computation. Our protocol privately executes the DBSCAN algorithm without disclosing any information (including the number and size of clusters). It can be used for private clustering between two parties as well as for private outsourcing of an arbitrary number of data owners to two non-colluding servers. Our implementation of the DBSCAN algorithm privately clusters data sets with 400 elements in 7 minutes on commodity hardware. Thereby, it flexibly determines the number of required clusters and is insensitive to outliers, while being only factor 19x slower than today's fastest private K-means protocol (Mohassel et al., PETS'20) which can only be used for specific data sets. We then show how to transfer our newly designed protocol to related clustering algorithms by introducing a private approximation of the TRACLUS algorithm for trajectory clustering which has interesting real-world applications like financial time series forecasts and the investigation of the spread of a disease like COVID-19.

Recent studies have shown that federated learning (FL) is vulnerable to poisoning attacks that inject a backdoor into the global model. These attacks are effective even when performed by a single client, and undetectable by most existing defensive techniques. In this paper, we propose Backdoor detection via Feedback-based Federated Learning (BAFFLE), a novel defense to secure FL against backdoor attacks. The core idea behind BAFFLE is to leverage data of multiple clients not only for training but also for uncovering model poisoning. We exploit the availability of diverse datasets at the various clients by incorporating a feedback loop into the FL process, to integrate the views of those clients when deciding whether a given model update is genuine or not. We show that this powerful construct can achieve very high detection rates against state-of-the-art backdoor attacks, even when relying on straightforward methods to validate the model. Through empirical evaluation using the CIFAR-10 and FEMNIST datasets, we show that by combining the feedback loop with a method that suspects poisoning attempts by assessing the per-class classification performance of the updated model, BAFFLE reliably detects state-of-the-art backdoor attacks with a detection accuracy of 100% and a false-positive rate below 5%. Moreover, we show that our solution can detect adaptive attacks aimed at bypassing the defense.

Abstract Clustering is a popular unsupervised machine learning technique that groups similar input elements into clusters. It is used in many areas ranging from business analysis to health care. In many of these applications, sensitive information is clustered that should not be leaked. Moreover, nowadays it is often required to combine data from multiple sources to increase the quality of the analysis as well as to outsource complex computation to powerful cloud servers. This calls for efficient privacy-preserving clustering. In this work, we systematically analyze the state-of-the-art in privacy-preserving clustering. We implement and benchmark today’s four most efficient fully private clustering protocols by Cheon et al. (SAC’19), Meng et al. (ArXiv’19), Mohassel et al. (PETS’20), and Bozdemir et al. (ASIACCS’21) with respect to communication, computation, and clustering quality. We compare them, assess their limitations for a practical use in real-world applications, and conclude with open challenges.