Techreport: Time-sensitive probabilistic inference for the edge

TL;DR

This paper explores probabilistic inference on edge computing systems by modeling time as a random variable within probabilistic programming, addressing the challenges of distributed data analysis.

Contribution

It introduces a formalism for time-sensitive probabilistic inference tailored for edge computing environments, integrating time as a stochastic element.

Findings

Formalism for probabilistic inference on the edge

Incorporation of time as a random variable

Potential for improved distributed data analysis

Abstract

In recent years the two trends of edge computing and artificial intelligence became both crucial for information processing infrastructures. While the centralized analysis of massive amounts of data seems to be at odds with computation on the outer edge of distributed systems, we explore the properties of eventually consistent systems and statistics to identify sound formalisms for probabilistic inference on the edge. In particular we treat time itself as a random variable that we incorporate into statistical models through probabilistic programming.