On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals
This work addresses the problem of optimizing sampling for graph-based classification tasks, which is incremental as it builds on existing graph signal processing methods.
The paper tackled the problem of graph signal reconstruction and sample selection for classification tasks by providing theoretical characterizations of classification error and comparing it to classical reconstruction error. The result was the derivation of new optimal sampling methods for linearized graph convolutional networks, showing improvement over other graph signal processing methods.
We investigate graph signal reconstruction and sample selection for classification tasks. We present general theoretical characterisations of classification error applicable to multiple commonly used reconstruction methods, and compare that to the classical reconstruction error. We demonstrate the applicability of our results by using them to derive new optimal sampling methods for linearized graph convolutional networks, and show improvement over other graph signal processing based methods.