Belief Propagation/Concurrent ML Venn diagram
figureposted on 2020-11-14, 03:45 authored by James CooperJames Cooper
A simple Venn diagram to illustrate the overlap between two of the major foci for my doctoral dissertation. Specifically, the topics of Belief Propagation and Concurrent ML, which are part of Computer Vision and Programming Languages, respectively.
Belief Propagation is an algorithm for computing inference over factor graphs, Markov Random Fields and similar. It is explicitly based around ideas of elements in the graph passing messages between themselves. It has been applied to the task of stereo matching (amongst others) in Computer Vision.
Concurrent ML is an approach to concurrent programming (and later also parallel programming) which is also explicitly based around ideas of computational elements passing messages between themselves.
Thus, the two seem like a natural fit, but no prior work on applying Concurrent ML to Belief Propagation has been found, leading to my dissertation topic. This diagram is a simple visual representation of the overlap/cross-over between the fields.