12. Červenec 2020
Understanding the robustness of Graph Neural Networks is crucial since they are widely used, yet highly sensitive to adversarial attacks. However, obtaining theoretical guarantees has been difficult so far due to the discrete and non-i.i.d. nature of graph data. Existing certificates handle either the graph structure or node attributes, but not both, and only work for a small class of models. We propose a randomized smoothing technique that overcomes these issues and furthermore generalizes previous certificates for binary data. Our approach explicitly accounts for sparsity in the input which, as our findings show, is essential for obtaining non-trivial guarantees. Moreover, our certificate is efficient and does not depend on the size of the input (e.g. the graph). We demonstrate the effectiveness of our approach on a wide variety of models, datasets, and tasks.
The International Conference on Machine Learning (ICML) is the premier gathering of professionals dedicated to the advancement of the branch of artificial intelligence known as machine learning. ICML is globally renowned for presenting and publishing cutting-edge research on all aspects of machine learning used in closely related areas like artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, and robotics. ICML is one of the fastest growing artificial intelligence conferences in the world. Participants at ICML span a wide range of backgrounds, from academic and industrial researchers, to entrepreneurs and engineers, to graduate students and postdocs.
Pro uložení prezentace do věčného trezoru hlasovalo 0 diváků, což je 0.0 %
Prezentace na podobné téma, kategorii nebo přednášejícího