Deep Reinforcement Learning-based Malicious URL Detection with Feature Selection.
Antonio MaciNicola TammaAntonio CosciaPublished in: ICAIC (2024)
Keyphrases
- reinforcement learning
- feature selection
- malicious url detection
- online learning
- cost sensitive
- machine learning
- function approximation
- text categorization
- state space
- multi class
- feature space
- reinforcement learning algorithms
- information gain
- unsupervised learning
- model selection
- model free
- mutual information
- text classification
- support vector machine
- feature extraction
- multi agent
- feature set
- small sample
- learning process
- selected features
- unsupervised feature selection
- feature selection algorithms
- optimal control
- optimal policy
- classification accuracy
- decision trees
- classification models
- multi task
- markov decision processes
- dimensionality reduction
- text mining
- temporal difference
- feature weighting
- supervised learning
- deep learning
- action space
- irrelevant features
- dynamic programming