Optimization and Interpretability of Graph Attention Networks for Small Sparse Graph Structures in Automotive Applications

Autor:	Neumeier, Marion, Tollkühn, Andreas, Dorn, Sebastian, Botsch, Michael, Utschick, Wolfgang
Rok vydání:	2023
Předmět:	Computer Science - Machine Learning Computer Science - Artificial Intelligence
Druh dokumentu:	Working Paper
Popis:	For automotive applications, the Graph Attention Network (GAT) is a prominently used architecture to include relational information of a traffic scenario during feature embedding. As shown in this work, however, one of the most popular GAT realizations, namely GATv2, has potential pitfalls that hinder an optimal parameter learning. Especially for small and sparse graph structures a proper optimization is problematic. To surpass limitations, this work proposes architectural modifications of GATv2. In controlled experiments, it is shown that the proposed model adaptions improve prediction performance in a node-level regression task and make it more robust to parameter initialization. This work aims for a better understanding of the attention mechanism and analyzes its interpretability of identifying causal importance. Comment: Accepted as a conference paper in IEEE IV 2023, Anchorage, Alaska, USA
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2305.16196 Zobrazit plný text záznamu View this record from Arxiv