Exchangeability-Aware Sum-Product Networks
This work addresses the need for efficient probabilistic modeling in relational domains with exchangeable variables, representing an incremental advancement by integrating existing model classes.
The paper tackled the problem of modeling exchangeable random variables in probabilistic models by introducing Exchangeability-Aware Sum-Product Networks (XSPNs), which combine SPNs and MEVMs to handle interchangeable data parts, resulting in improved accuracy over conventional SPNs as shown empirically.
Sum-Product Networks (SPNs) are expressive probabilistic models that provide exact, tractable inference. They achieve this efficiency by making use of local independence. On the other hand, mixtures of exchangeable variable models (MEVMs) are a class of tractable probabilistic models that make use of exchangeability of discrete random variables to render inference tractable. Exchangeability, which arises naturally in relational domains, has not been considered for efficient representation and inference in SPNs yet. The contribution of this paper is a novel probabilistic model which we call Exchangeability-Aware Sum-Product Networks (XSPNs). It contains both SPNs and MEVMs as special cases, and combines the ability of SPNs to efficiently learn deep probabilistic models with the ability of MEVMs to efficiently handle exchangeable random variables. We introduce a structure learning algorithm for XSPNs and empirically show that they can be more accurate than conventional SPNs when the data contains repeated, interchangeable parts.