Hierarchical graph embedding in vector space by graph pyramid

作者：

Highlights：

• We propose a hierarchical framework which extracts local and global features from different scales of given graph at the same time.

• Reducing the computational complexity with processing at higher levels of abstraction.

• Independent processing of different resolutions of graph based on application.

• Bridging the gap between elementary features and more descriptive ones.

摘要

Highlights•We propose a hierarchical framework which extracts local and global features from different scales of given graph at the same time.•Reducing the computational complexity with processing at higher levels of abstraction.•Independent processing of different resolutions of graph based on application.•Bridging the gap between elementary features and more descriptive ones.

论文关键词：Graph embedding,Multi-resolution theory,Graph pyramid,Graph summarization,Graph classification

论文评审过程：Received 14 October 2015, Revised 13 May 2016, Accepted 30 July 2016, Available online 1 August 2016, Version of Record 10 August 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.07.043