Recurrent Neural Network Language Models for Open Vocabulary Event-Level Cyber Anomaly Detection.
Aaron Randall TuorRyan BaerwolfNicolas KnowlesBrian HutchinsonNicole NicholsRobert JasperPublished in: AAAI Workshops (2018)
Keyphrases
- anomaly detection
- language model
- recurrent neural networks
- language modeling
- intrusion detection
- speech recognition
- n gram
- detecting anomalies
- information retrieval
- spoken term detection
- document retrieval
- network intrusion detection
- neural network
- probabilistic model
- out of vocabulary
- statistical language models
- one class support vector machines
- retrieval model
- anomalous behavior
- network traffic
- network anomaly detection
- unsupervised learning
- query expansion
- recurrent networks
- feed forward
- intrusion detection system
- query terms
- artificial neural networks
- language models for information retrieval
- machine learning
- novelty detection
- support vector
- feature selection
- relevance model
- back propagation
- smoothing methods
- abnormal events
- text mining
- object recognition