Convolutional Learning on Directed Acyclic Graphs

TL;DR

This paper introduces a new convolutional neural network architecture designed specifically for data on directed acyclic graphs, addressing unique challenges posed by DAG structures in signal processing and machine learning.

Contribution

It proposes a novel DAG convolutional neural network that incorporates learnable DAG filters to effectively model the partial ordering in DAGs, enhancing representation learning.

Findings

DCN outperforms several baseline methods on synthetic tasks

The approach effectively captures causal relationships in DAG-structured data

Demonstrates promising potential for DAG-based learning applications

Abstract

We develop a novel convolutional architecture tailored for learning from data defined over directed acyclic graphs (DAGs). DAGs can be used to model causal relationships among variables, but their nilpotent adjacency matrices pose unique challenges towards developing DAG signal processing and machine learning tools. To address this limitation, we harness recent advances offering alternative definitions of causal shifts and convolutions for signals on DAGs. We develop a novel convolutional graph neural network that integrates learnable DAG filters to account for the partial ordering induced by the graph topology, thus providing valuable inductive bias to learn effective representations of DAG-supported data. We discuss the salient advantages and potential limitations of the proposed DAG convolutional network (DCN) and evaluate its performance on two learning tasks using synthetic data: network diffusion estimation and source identification. DCN compares favorably relative to several baselines, showcasing its promising potential.