Divide and Rule: Recurrent Partitioned Network for Dynamic Processes

TL;DR

This paper introduces REIN, a novel recurrent partitioned network that models dynamic systems by capturing hierarchical interactions and causal dependencies, improving long-term predictions with limited data.

Contribution

The paper presents a new architecture combining hierarchical representation, relational inference, and statistical prediction for dynamic systems modeling.

Findings

Effective in identifying component interactions with limited data

Stable long-term future predictions across diverse physical systems

Outperforms existing methods in capturing intra-system dependencies

Abstract

In general, many dynamic processes are involved with interacting variables, from physical systems to sociological analysis. The interplay of components in the system can give rise to confounding dynamic behavior. Many approaches model temporal sequences holistically ignoring the internal interaction which are impotent in capturing the protogenic actuation. Differently, our goal is to represent a system with a part-whole hierarchy and discover the implied dependencies among intra-system variables: inferring the interactions that possess causal effects on the sub-system behavior with REcurrent partItioned Network (REIN). The proposed architecture consists of (i) a perceptive module that extracts a hierarchical and temporally consistent representation of the observation at multiple levels, (ii) a deductive module for determining the relational connection between neurons at each level, and (iii) a statistical module that can predict the future by conditioning on the temporal distributional estimation. Our model is demonstrated to be effective in identifying the componential interactions with limited observation and stable in long-term future predictions experimented with diverse physical systems.