Privacy accounting and quality control in the sage differentially private ML platform

Citations Over TimeTop 10% of 2019 papers

Abstract

Companies increasingly expose machine learning (ML) models trained over sensitive user data to untrusted domains, such as end-user devices and wide-access model stores. This creates a need to control the data's leakage through these models. We present Sage, a differentially private (DP) ML platform that bounds the cumulative leakage of training data through models. Sage builds upon the rich literature on DP ML algorithms and contributes pragmatic solutions to two of the most pressing systems challenges of global DP: running out of privacy budget and the privacy-utility tradeoff. To address the former, we develop block composition, a new privacy loss accounting method that leverages the growing database regime of ML workloads to keep training models endlessly on a sensitive data stream while enforcing a global DP guarantee for the stream. To address the latter, we develop privacy-adaptive training, a process that trains a model on growing amounts of data and/or with increasing privacy parameters until, with high probability, the model meets developer-configured quality criteria. Sage's methods are designed to integrate with TensorFlow-Extended, Google's open-source ML platform. They illustrate how a systems focus on characteristics of ML workloads enables pragmatic solutions that are not apparent when one focuses on individual algorithms, as most DP ML literature does.

Related Papers

• → Chemical and Genetic Relationships among Sage (Salvia officinalis L.) Cultivars and Judean Sage (Salvia judaica Boiss.)(2009)59 cited
• → Antimutagenic effect of sage tea in the wing spot test of Drosophila melanogaster(2008)37 cited
• → EFFECT OF DIRECT APPLICATIONS OF SAGE (SALVIA OFFICINALIS L.) LEAVES ON OXIDATIVE STABILITY OF SUNFLOWER OIL DURING ACCELERATED STORAGE(2009)5 cited
• SAGE and its applications(2002)
• → CHAPTER 25. Sage/Common Sage (Salvia officinalis)(2021)