Atlas convolutional neural network data processing method and device

Abstract

The embodiment of the invention provides a graph convolutional neural network data processing method and device. The method comprises the steps: carrying out the clustering and merging of all nodes ina graph, and obtaining a coarsened graph; filtering the unstructured atlas in the convolutional neural network according to a preset convolutional filter to obtain a filtered convolutional neural network; according to the method, the limitation of non-parameterization in high-dimensional data application can be effectively solved, the application generalization ability is improved, the calculation complexity is reduced, and the operation efficiency is improved.

Description

technical field [0001] The present application relates to the field of data processing, in particular to a data processing method and device for a graph convolutional neural network. Background technique [0002] Graph or network widely exists in daily life, and it is an important data structure for abstracting the relationship between objects in the real world. Data such as the citation relationship between authors, the social relationship between individuals, the logistics and transportation relationship between cities, and the interaction relationship between proteins can be abstractly expressed through graphs or networks. The analysis and modeling of such data can mine rich potential information, which can be widely used in node classification, community discovery, link prediction, recommendation system and other tasks. [0003] Traditional network representations such as adjacency matrices suffer from sparse structure and high dimensionality, making them difficult to l...

AI Technical Summary

Problems solved by technology

The usual convolutional neural network cannot directly use the convolutional layer and the p

Images