Simulating SQL Injection Vulnerability Exploitation Using Q-Learning Reinforcement Learning Agents.
Laszlo ErdodiÅvald Åslaugson SommervollFabio Massimo ZennaroPublished in: CoRR (2021)
Keyphrases
- reinforcement learning agents
- reinforcement learning
- security vulnerabilities
- sql injection
- multi agent
- state abstraction
- state space
- function approximation
- unauthorized access
- reinforcement learning algorithms
- dynamic environments
- learning algorithm
- injection attacks
- web applications
- intrusion detection
- markov decision processes
- model free
- multi agent environments
- optimal policy
- cooperative
- action selection
- transfer learning
- dynamic programming
- database
- security issues
- single agent
- supervised learning
- access control
- machine learning
- reward function
- dos attacks
- multiple agents
- countermeasures
- reinforcement learning methods
- multi agent systems