Non-Local Feature Aggregation on Graphs via Latent Fixed Data Structures
This addresses a bottleneck in GNNs for tasks requiring non-local feature aggregation, offering a novel method that is incremental in improving existing approaches.
The paper tackles the problem of global feature aggregation in Graph Neural Networks (GNNs) by introducing a Latent Fixed Data Structure (LFDS) to sort and distribute locally extracted feature vectors, enabling non-local aggregation via latent neural networks like CNNs or GNNs. It achieves competitive or better results with linear computational complexity relative to input graph order.
In contrast to image/text data whose order can be used to perform non-local feature aggregation in a straightforward way using the pooling layers, graphs lack the tensor representation and mostly the element-wise max/mean function is utilized to aggregate the locally extracted feature vectors. In this paper, we present a novel approach for global feature aggregation in Graph Neural Networks (GNNs) which utilizes a Latent Fixed Data Structure (LFDS) to aggregate the extracted feature vectors. The locally extracted feature vectors are sorted/distributed on the LFDS and a latent neural network (CNN/GNN) is utilized to perform feature aggregation on the LFDS. The proposed approach is used to design several novel global feature aggregation methods based on the choice of the LFDS. We introduce multiple LFDSs including loop, 3D tensor (image), sequence, data driven graphs and an algorithm which sorts/distributes the extracted local feature vectors on the LFDS. While the computational complexity of the proposed methods are linear with the order of input graphs, they achieve competitive or better results.