Approximation enhancement for stochastic Bayesian inference,☆☆

https://doi.org/10.1016/j.ijar.2017.03.007Get rights and content
Under an Elsevier user license
open archive

Highlights

  • Orders-of-magnitude improvement in approximate Bayesian inference efficiency.

  • Bitstream autocorrelation limits inference approximation accuracy.

  • Autocorrelation successfully mitigated to improve Bayesian inference approximation.

  • Approximate Bayesian inference efficiently performed in hardware.

Abstract

Advancements in autonomous robotic systems have been impeded by the lack of a specialized computational hardware that makes real-time decisions based on sensory inputs. We have developed a novel circuit structure that efficiently approximates naïve Bayesian inference with simple Muller C-elements. Using a stochastic computing paradigm, this system enables real-time approximate decision-making with an area-energy-delay product nearly one billion times smaller than a conventional general-purpose computer. In this paper, we propose several techniques to improve the approximation of Bayesian inference by reducing stochastic bitstream autocorrelation. We also evaluate the effectiveness of these techniques for various naïve inference tasks and discuss hardware considerations, concluding that these circuits enable approximate Bayesian inferences while retaining orders-of-magnitude hardware advantages compared to conventional general-purpose computers.

Keywords

Stochastic computing
Muller C-element
Bayesian inference
Autocorrelation
Approximate inference

Cited by (0)

This work was supported by the EU collaborative FET Project BAMBI FP7-ICT-2013-C, project number 618024 and a public grant overseen by the French National Research Agency (ANR) as part of the “Investissements d'Avenir” program (Labex NanoSaclay, reference: ANR-10-LABX-0035).

☆☆

This paper is part of the Virtual special issue on Special Issue on Unconventional computing for Bayesian inference, edited by Jorge Lobo and João Filipe Ferreira.