Extreme Learning Machine for large-scale graph classification based on MapReduce

Abstract

Discriminative subgraph mining from a large collection of graph objects is a crucial problem for graph classification. Several main memory-based approaches have been proposed to mine discriminative subgraphs, but they always lack scalability and are not suitable for large-scale graph databases. Extreme Learning Machine (ELM) is a simple and efficient Single-hidden Layer Feedforward neural Networks (SLFNs) algorithm with extremely fast learning capacity. In this paper, we propose a discriminative subgraph mining approach based on ELM-Filter strategy within the scalable MapReduce computing model. We randomly partition the collection of graphs among worker nodes, and each worker applies a fast pattern evolutionary method to mine a set of discriminative subgraphs with the help of ELM-Filter strategy in its partition. And, the set of discriminative subgraphs must produce higher ELM training accuracy. The union of all such discriminative subgraphs is the mining result for the input large-scale graphs. Also, based on the proposed Support Graph Vector Model (SGVM) and ELM algorithm, we construct the graph classification model using the mined discriminative subgraphs. Extensive experimental results on both real and synthetic datasets show that our method obviously outperforms the other approaches in terms of both classification accuracy and runtime efficiency.