An Information-theoretic Approach for Privacy Preservation in Distance-based Machine Learning
As cloud-based services become increasingly popular as platforms for storage and computation, privacy issues relating to their use have become increasingly important. Much of the data stored on cloud platforms are private, belonging to individuals or institutions who often desire to utilize the facilities provided by these platforms, but, at the same time, do not desire to expose their data to the platform itself.
Encrypting the data prior to storage on the cloud helps to protect private information. However, this causes problems if we need to perform computations on them, for instance, to train some machine learning algorithm. This requires the server to observe the content, so decryption is necessary. This gives rise to privacy concerns in different cloud computing settings. Several solutions based on cryptographic techniques have been proposed to address the issue. However, they have high computational cost and high bandwidth requirements, and in practice are difficult to scale.
In this talk, we will propose an alternative approach. We introduce a new privacy mechanism, named limited leakage transformation, that employs custom transformations based on information-theoretic principles, which both hide data, and also permit limited computation of metrics required for machine learning and inference, with little to no computational overhead, thus potentially enabling practical large-scale privacy-preserving ML. We state information-theoretic properties that allow us to analyze our privacy guarantees and describe how to use this kind of scheme in practical scenarios. We will discuss the impact of this approach in different applications, such as Private Image Retrieval and Cancelable Biometrics, among others.
Friday, June 28 at 10:00 a.m.