How Wrong Am I? - Studying Adversarial Examples and their Impact on Uncertainty in Gaussian Process Machine Learning Models
Explore this paper's citation graph
Summary
Gaussian Processes is used to investigate adversarial examples in the framework of Bayesian inference and finds deviating levels of uncertainty reflect the perturbation introduced to benign samples by state-of-the-art attacks, including novel white-box attacks on Gaussian Processe.
- Type
- preprint
- Published
- 2017-11-17
- Cited by
- 9
- References
- 60
- Access
- Open access
- OpenAlex
- https://openalex.org/W2770947558
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:3363528
Keywords
Adversarial system, Gaussian process, Process (computing), Artificial intelligence, Computer science
References
- Good Word Attacks on Statistical Spam Filters
- Statistical Meta-Analysis of Presentation Attacks for Secure Multibiometric Systems
- Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning)
- Using the Taylor expansion of multilayer feedforward neural networks
- Intriguing properties of neural networks
- Automatic analysis of malware behavior using machine learning
- Deep neural network based malware detection using two dimensional binary program features
- Outside the Closed World: On Using Machine Learning for Network Intrusion Detection
- Looking at the bag is not enough to find the bomb: an evasion of structural methods for malicious PDF files detection
- Practical Evasion of a Learning-Based Classifier: A Case Study
- Gradient-based learning applied to document recognition
- An evaluation of Naive Bayesian anti-spam filtering
- The security of machine learning
- Scene parsing with Multiscale Feature Learning, Purity Trees, and Optimal Covers
- Gaussian Process Latent Variable Models for Visualisation of High Dimensional Data
- The Limitations of Deep Learning in Adversarial Settings
- DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks
- Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples
- Curie: A method for protecting SVM Classifier from Poisoning Attack
- Adversarial Perturbations Against Deep Neural Networks for Malware Classification
Cited by
- Killing Three Birds with one Gaussian Process: Analyzing Attack Vectors on Classification
- Bayesian deep learning on a quantum computer
- Robustness Guarantees for Bayesian Inference with Gaussian Processes
- Why the Failure? How Adversarial Examples Can Provide Insights for Interpretable Machine Learning
- Quantum Statistical Inference
- Defence against adversarial attacks using classical and quantum-enhanced Boltzmann machines
- Probabilistic Reach-Avoid for Bayesian Neural Networks
- Adversarial Robustness Guarantees for Gaussian Processes
- Certification of Iterative Predictions in Bayesian Neural Networks