Sparse Representation Classification via Screening for Graphs
This work addresses computational bottlenecks in graph classification for researchers and practitioners, though it is incremental as it builds on existing SRC methods.
The authors tackled the problem of improving the computational efficiency of the sparse representation classifier (SRC) for graph data by proposing a new implementation via screening, which achieved comparable classification performance but significantly faster speeds in simulations and real data experiments.
The sparse representation classifier (SRC) is shown to work well for image recognition problems that satisfy a subspace assumption. In this paper we propose a new implementation of SRC via screening, establish its equivalence to the original SRC under regularity conditions, and prove its classification consistency for random graphs drawn from stochastic blockmodels. The results are demonstrated via simulations and real data experiments, where the new algorithm achieves comparable numerical performance but significantly faster.