arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.14282cs.LGcs.AI
收藏

从消息传递重新思考图神经结构搜索

Rethinking Graph Neural Architecture Search from Message-passing

University of Chinese Academy of Sciences(中国科学院大学);University of Science and Technology of China(中国科学技术大学);Peng Cheng Laboratory(鹏城实验室)

Shaofei Cai, Liang Li, Jincan Deng, Beichen Zhang, Zheng-Jun Zha, Li Su, Qingming Huang

摘要:

图神经网络(GNNs)最近作为从图上的数据学习的标准工具包出现。目前的GNN设计工作依赖于大量的人类专业知识来探索不同的消息传递机制,并需要手动枚举来确定适当的消息传递深度。受CNN中神经结构搜索(NAS)强大搜索能力的启发,提出了一种新的搜索空间的图神经结构搜索(GNAS)。GNAS可以自动学习更好的架构,并在图上获得最佳的消息传递深度。具体来说,我们设计了图神经架构范式(GAP),它采用树型拓扑计算过程和两种细粒度原子操作(特征过滤和邻居聚合),从消息传递机制出发,构建了强大的图网络搜索空间。特征过滤执行自适应特征选择,而邻居聚合捕获结构信息并计算邻居的统计信息。实验表明,我们的GNAS可以搜索更好的GNN与多种消息传递机制和最佳的消息传递深度。搜索网络在三个经典图任务的五个大规模数据集上比最先进的手动设计和基于搜索的GNN取得了显着的改进。代码可以在https://github.com/phython96/GNAS-MP上找到。

英文摘要:

Graph neural networks (GNNs) emerged recently as a standard toolkit for learning from data on graphs. Current GNN designing works depend on immense human expertise to explore different message-passing mechanisms, and require manual enumeration to determine the proper message-passing depth. Inspired by the strong searching capability of neural architecture search (NAS) in CNN, this paper proposes Graph Neural Architecture Search (GNAS) with novel-designed search space. The GNAS can automatically learn better architecture with the optimal depth of message passing on the graph. Specifically, we design Graph Neural Architecture Paradigm (GAP) with tree-topology computation procedure and two types of fine-grained atomic operations (feature filtering and neighbor aggregation) from message-passing mechanism to construct powerful graph network search space. Feature filtering performs adaptive feature selection, and neighbor aggregation captures structural information and calculates neighbors' statistics. Experiments show that our GNAS can search for better GNNs with multiple message-passing mechanisms and optimal message-passing depth. The searched network achieves remarkable improvement over state-of-the-art manual designed and search-based GNNs on five large-scale datasets at three classical graph tasks. Codes can be found at https://github.com/phython96/GNAS-MP.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑