No Time to Lie: Bounds on the Learning Rate of a Defender for Inferring Attacker Target Preferences.
Mark BilinskiJoseph DiVitaKimberly Ferguson-WalterSunny FugateRyan GabrysJustin MaugerBrian SouzaPublished in: GameSec (2021)
Keyphrases
- learning rate
- uniform convergence
- covering numbers
- learning algorithm
- convergence rate
- error function
- hidden layer
- lower bound
- upper bound
- convergence speed
- rapid convergence
- multilayer neural networks
- worst case
- adaptive learning rate
- delta bar delta
- training algorithm
- game theory
- vc dimension
- training data
- neural network
- countermeasures
- convergence theorem