Network Signatures from Image Representation of Adjacency Matrices: Deep/Transfer Learning for Subgraph Classification
This addresses subgraph classification for network analysis, offering an incremental improvement in representation and learning methods.
The paper tackled subgraph classification by representing adjacency matrices as 2D images, using deep learning and transfer learning; results showed deep learning with these features outperformed benchmarks and transfer learning was effective with minimal data.
We propose a novel subgraph image representation for classification of network fragments with the targets being their parent networks. The graph image representation is based on 2D image embeddings of adjacency matrices. We use this image representation in two modes. First, as the input to a machine learning algorithm. Second, as the input to a pure transfer learner. Our conclusions from several datasets are that (a) deep learning using our structured image features performs the best compared to benchmark graph kernel and classical features based methods; and, (b) pure transfer learning works effectively with minimum interference from the user and is robust against small data.