1: \begin{abstract}%

2:   We introduce a new type of graphical model that we call a 

3:   ``memory factor network''

4:   (MFN).  We show how to use MFNs to model the structure inherent in 

5:   many types of data sets.  We also introduce an associated

6:   message-passing style algorithm called ``proactive message passing'' (PMP)

7:   that performs inference on MFNs. PMP comes with

8:   convergence guarantees and is efficient in comparison to

9:   competing algorithms such as variants of belief propagation.  We

10:   specialize MFNs and PMP to a number of distinct types of data

11:   (discrete, continuous, labelled) and inference problems

12:   (interpolation, hypothesis testing), provide examples, and discuss

13:   approaches for efficient implementation.

14: \end{abstract}

15: