Online Variance Reduction with Mixtures [article]

Zalán Borsos, Sebastian Curi, Kfir Y. Levy, Andreas Krause
2019 arXiv   pre-print
Adaptive importance sampling for stochastic optimization is a promising approach that offers improved convergence through variance reduction. In this work, we propose a new framework for variance reduction that enables the use of mixtures over predefined sampling distributions, which can naturally encode prior knowledge about the data. While these sampling distributions are fixed, the mixture weights are adapted during the optimization process. We propose VRM, a novel and efficient adaptive
more » ... me that asymptotically recovers the best mixture weights in hindsight and can also accommodate sampling distributions over sets of points. We empirically demonstrate the versatility of VRM in a range of applications.
arXiv:1903.12416v1 fatcat:avpyaq5v2fa3xavqdx65c7o4je