Timezone: »

 
Differentially Private Adaptive Optimization with Delayed Preconditioners
Tian Li · Manzil Zaheer · Ken Liu · Sashank Reddi · H. Brendan McMahan · Virginia Smith

@
in OPT 2022: Optimization for Machine Learning »
Event URL: https://openreview.net/forum?id=jKnTWwW8eVN »

Privacy costs may negate the benefits of using adaptive optimizers in differentially private model training. Prior works typically address this issue by using auxiliary information (e.g., public data) to boost the effectiveness of adaptive optimization. In this work, we explore techniques to estimate and efficiently adapt to gradient geometry in private adaptive optimization without auxiliary data. Motivated by the observation that adaptive methods can tolerate stale preconditioners, we propose differentially private adaptive training with delayed preconditioners (DP^2), a simple method that constructs delayed but less noisy preconditioners to better realize the benefits of adaptivity. Theoretically, we provide convergence guarantees for our method for both convex and non-convex problems, and analyze trade-offs between delay and privacy noise reduction. Empirically, we explore DP^2 across several real-world datasets, demonstrating that it can improve convergence speed by as much as 4× relative to non-adaptive baselines and match the performance of state-of-the-art optimization methods that require auxiliary data.

Author Information

Tian Li (CMU)
Manzil Zaheer (Google)
Ken Liu (Carnegie Mellon University)
Sashank Reddi (Google)
H. Brendan McMahan (Google, Inc.)
Virginia Smith (Carnegie Mellon University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors