New efficient neural network approach for processing graph structured data

Abstract

Graphs serve as important data structures in many applications including so newlinecial network web and Bio informatics Recursive Neural NetworksRNN are a newlineconnectionist model designed to process structured data RNN are particularly newlinesuited to process Directed Positional Acyclic GraphsDPAGs Graph Neural Net newlineworkGNN has been modeled to process data represented in graph domains con newlinesidering the topology of the graph newline newline