Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Streaming Bayesian inference: theoretical limits and mini-batch approximate message-passing

Abstract : In statistical learning for real-world large-scale data problems, one must often resort to "streaming" algorithms which operate sequentially on small batches of data. In this work, we present an analysis of the information-theoretic limits of mini-batch inference in the context of generalized linear models and low-rank matrix factorization. In a controlled Bayes-optimal setting, we characterize the optimal performance and phase transitions as a function of mini-batch size. We base part of our results on a detailed analysis of a mini-batch version of the approximate message-passing algorithm (Mini-AMP), which we introduce. Additionally, we show that this theoretical optimality carries over into real-data problems by illustrating that Mini-AMP is competitive with standard streaming algorithms for clustering.
Complete list of metadata

Cited literature [43 references]  Display  Hide  Download
Contributor : Emmanuelle De Laborderie Connect in order to contact the contributor
Submitted on : Monday, July 3, 2017 - 4:04:57 PM
Last modification on : Thursday, March 17, 2022 - 10:08:10 AM


Files produced by the author(s)


  • HAL Id : cea-01553517, version 1
  • ARXIV : 1706.00705


Andre Manoel, Florent Krzakala, Eric W. Tramel, Lenka Zdeborová. Streaming Bayesian inference: theoretical limits and mini-batch approximate message-passing. 2017. ⟨cea-01553517⟩



Record views


Files downloads