Adversarial Machine Learning Phases of Matter
arXiv:1910.13453 · doi:10.1007/s44214-023-00043-z
Abstract
We study the robustness of machine learning approaches to adversarial perturbations, with a focus on supervised learning scenarios. We find that typical phase classifiers based on deep neural networks are extremely vulnerable to adversarial perturbations: adding a tiny amount of carefully crafted noises into the original legitimate examples will cause the classifiers to make incorrect predictions at a notably high confidence level. Through the lens of activation maps, we find that some important underlying physical principles and symmetries remain to be adequately captured for classifiers with even near-perfect performance. This explains why adversarial perturbations exist for fooling these classifiers. In addition, we find that, after adversarial training, the classifiers will become more consistent with physical laws and consequently more robust to certain kinds of adversarial perturbations. Our results provide valuable guidance for both theoretical and experimental future studies on applying machine learning techniques to condensed matter physics.
5 pages main text plus 9 pages supplementary materials
References in corpus (20)
- Explaining and Harnessing Adversarial Examples
- Learning phase transitions by confusion
- Object Detectors Emerge in Deep Scene CNNs
- Discovering Phases, Phase Transitions and Crossovers through Unsupervised Machine Learning: A critical examination
- Restricted-Boltzmann-Machine Learning for Solving Strongly Correlated Quantum Systems
- Unsupervised machine learning and band topology
- Machine learning meets quantum physics
- Deep Learning the Quantum Phase Transitions in Random Two-Dimensional Electron Systems
- Unsupervised Manifold Clustering of Topological Phononics
- Topological quantum phase transitions retrieved through unsupervised machine learning
- Deep Learning the Quantum Phase Transitions in Random Electron Systems: Applications to Three Dimensions
- Machine Learning Topological Phases with a Solid-state Quantum Simulator
- Unsupervised identification of topological order using predictive models
- Unsupervised machine learning of quantum phase transitions using diffusion maps
- Unsupervised Learning of Non-Hermitian Topological Phases
- Drawing Phase Diagrams of Random Quantum Systems by Deep Learning the Wave Functions
- Unsupervised learning using topological data augmentation
- Interpretable and unsupervised phase classification
- Visualizing Quantum Phases And Identifying Quantum Phase Transitions By Nonlinear Dimensionality Reduction
- Featuring the topology with the unsupervised machine learning