Promise and Limitations of Supervised Optimal Transport-Based Graph Summarization via Information Theoretic Measures

被引:2
|
作者
Neshatfar, Sepideh [1 ]
Magner, Abram [2 ]
Sekeh, Salimeh Yasaei [1 ]
机构
[1] Univ Maine, Sch Comp & Informat Sci, Orono, ME 04469 USA
[2] SUNY Albany, Coll Engn & Appl Sci, Albany, NY 12222 USA
基金
美国国家科学基金会;
关键词
Graph classification; monotonicity; optimal transport; Shannon mutual information; supervised graph summarization;
D O I
10.1109/ACCESS.2023.3302830
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Graph summarization is the problem of producing smaller graph representations of an input graph dataset, in such a way that the smaller compressed graphs capture relevant structural information for downstream tasks. One of the recent graph summarization methods formulates an optimal transport-based framework that allows prior information about node, edge, and attribute importance to be incorporated into the graph summarization process. However, very little is known about the statistical properties of this framework. To elucidate this question, we consider the problem of supervised graph summarization, wherein by using information theoretic measures we seek to preserve relevant information about a class label. To gain a theoretical perspective on the supervised summarization problem itself, we first formulate it in terms of maximizing the Shannon mutual information between the summarized graph and the class label. We show an NP-hardness of approximation result for this problem, thereby constraining what one should expect from proposed solutions. We then propose a summarization method that incorporates mutual information estimates between random variables associated with sample graphs and class labels into the optimal transport compression framework. We empirically show performance improvements over previous works in terms of classification accuracy and time on synthetic and certain real datasets. We also theoretically explore the limitations of the optimal transport approach for the supervised summarization problem and we show that it fails to satisfy a certain desirable information monotonicity property.
引用
收藏
页码:87533 / 87542
页数:10
相关论文
共 4 条
  • [1] OPTIMAL TRANSPORT-BASED GRAPH MATCHING FOR 3D RETINAL OCT IMAGE REGISTRATION
    Tian, Xin
    Anantrasirichai, Nantheera
    Nicholson, Lindsay
    Achim, Alin
    2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2022, : 2791 - 2795
  • [2] Enhancing robust semi-supervised graph alignment via adaptive optimal transport
    Chen, Songyang
    Lin, Youfang
    Liu, Yu
    Ouyang, Yuwei
    Guo, Zongshen
    Zou, Lei
    WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS, 2025, 28 (02):
  • [3] Enhancing Multi-modal Contrastive Learning via Optimal Transport-Based Consistent Modality Alignment
    Zhu, Sidan
    Luo, Dixin
    PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2024, PT XI, 2025, 15041 : 157 - 171
  • [4] Improving Geological Remote Sensing Interpretation via Optimal Transport-Based Point-Surface Data Fusion
    Wu, Jiahao
    Han, Wei
    Chen, Jia
    Wang, Sheng
    REMOTE SENSING, 2024, 16 (01)