The mathematics of adversarial attacks in AI - Why deep learning is unstable despite the existence of stable neural networks.
Alexander BastounisAnders C. HansenVerner VlacicPublished in: CoRR (2021)
Keyphrases
- deep learning
- neural network
- machine learning
- unsupervised learning
- artificial intelligence
- pattern recognition
- unsupervised feature learning
- expert systems
- weakly supervised
- deep architectures
- back propagation
- neural nets
- mental models
- restricted boltzmann machine
- face recognition
- knowledge representation
- decision making
- natural language
- semi supervised
- deep belief networks