| Title | Node Copying for Protection Against Graph Neural Network Topology Attacks |
| Publication Type | Conference Paper |
| Year of Publication | 2019 |
| Authors | Regol, Florence, Pal, Soumyasundar, Coates, Mark |
| Conference Name | 2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP) |
| Date Published | dec |
| Keywords | adversarial attacks, Attack Graphs, composability, Computational modeling, corruption, deep learning models, detection problem, downstream learning task, graph based machine, graph connectivity, graph convolutional networks, graph neural network topology attacks, graph theory, graph topology, learning (artificial intelligence), network theory (graphs), Network topology, neural nets, Neural networks, node copying, Prediction algorithms, prediction capability, Predictive Metrics, pubcrawl, Resiliency, security of data, semi-supervised learning, similarity structure, Task Analysis, Topology, Training |
| Abstract | Adversarial attacks can affect the performance of existing deep learning models. With the increased interest in graph based machine learning techniques, there have been investigations which suggest that these models are also vulnerable to attacks. In particular, corruptions of the graph topology can degrade the performance of graph based learning algorithms severely. This is due to the fact that the prediction capability of these algorithms relies mostly on the similarity structure imposed by the graph connectivity. Therefore, detecting the location of the corruption and correcting the induced errors becomes crucial. There has been some recent work which tackles the detection problem, however these methods do not address the effect of the attack on the downstream learning task. In this work, we propose an algorithm that uses node copying to mitigate the degradation in classification that is caused by adversarial attacks. The proposed methodology is applied only after the model for the downstream task is trained and the added computation cost scales well for large graphs. Experimental results show the effectiveness of our approach for several real world datasets. |
| DOI | 10.1109/CAMSAP45676.2019.9022508 |
| Citation Key | regol_node_2019 |
Node Copying for Protection Against Graph Neural Network Topology Attacks