Timezone: »

Variance Reduction in Stochastic Gradient Langevin Dynamics
Kumar Avinava Dubey · Sashank J. Reddi · Sinead Williamson · Barnabas Poczos · Alexander Smola · Eric Xing

Wed Dec 07 09:00 AM -- 12:30 PM (PST) @ Area 5+6+7+8 #9

Stochastic gradient-based Monte Carlo methods such as stochastic gradient Langevin dynamics are useful tools for posterior inference on large scale datasets in many machine learning applications. These methods scale to large datasets by using noisy gradients calculated using a mini-batch or subset of the dataset. However, the high variance inherent in these noisy gradients degrades performance and leads to slower mixing. In this paper, we present techniques for reducing variance in stochastic gradient Langevin dynamics, yielding novel stochastic Monte Carlo methods that improve performance by reducing the variance in the stochastic gradient. We show that our proposed method has better theoretical guarantees on convergence rate than stochastic Langevin dynamics. This is complemented by impressive empirical results obtained on a variety of real world datasets, and on four different machine learning tasks (regression, classification, independent component analysis and mixture modeling). These theoretical and empirical contributions combine to make a compelling case for using variance reduction in stochastic Monte Carlo methods.

Author Information

Kumar Avinava Dubey (Carnegie Mellon University)
Sashank J. Reddi (Carnegie Mellon University)
Sinead Williamson (University of Texas at Austin)
Barnabas Poczos (Carnegie Mellon University)
Alexander Smola (Amazon)

**AWS Machine Learning**

Eric Xing (Carnegie Mellon University)

More from the Same Authors