Recurrent Neural Network Attention Mechanisms for Interpretable System Log Anomaly Detection
Explore this paper's citation graph
Summary
Recurrent neural network language models augmented with attention for anomaly detection in system logs are presented, creating opportunities for model introspection and analysis without sacrificing state-of-the art performance.
- Type
- preprint
- Published
- 2018-03-13
- Cited by
- 198
- References
- 24
- Access
- Open access
- OpenAlex
- https://openalex.org/W2792645615
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:3827029
Keywords
Interpretability, Computer science, Artificial intelligence, Anomaly detection, Deep learning
References
- Making machine learning models interpretable
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- Malware classification with recurrent networks
- Learning to Transduce with Unbounded Memory
- Effective Approaches to Attention-based Neural Machine Translation
- Long Short-Term Memory
- Principal component analysis
- Reasoning about Entailment with Neural Attention
- The mythos of model interpretability
- The ethics of algorithms: Mapping the debate
- DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning
- Recurrent Neural Network Language Models for Open Vocabulary Event-Level Cyber Anomaly Detection
- Attentive Language Models
- Memory Architectures in Recurrent Neural Network Language Models
- Explaining hyperspectral imaging based plant disease identification: 3D CNN and saliency maps
- Attention is All you Need
- Automated IT system failure prediction: A deep learning approach
- Understanding intermediate layers using linear classifier probes
- Adam: A Method for Stochastic Optimization
- Neural Machine Translation by Jointly Learning to Align and Translate
Cited by
- Sequence Aggregation Rules for Anomaly Detection in Computer Network Traffic
- Attention Models in Graphs
- Large scale anomaly detection in data center logs and metrics
- Detecting Anomaly in Big Data System Logs Using Convolutional Neural Network
- DReAM: Deep Recursive Attentive Model for Anomaly Detection in Kernel Events
- Defending Our Public Biological Databases as a Global Critical Infrastructure
- Finding Rats in Cats: Detecting Stealthy Attacks using Group Anomaly Detection
- User-Assisted Log Analysis for Quality Control of Distributed Fintech Applications
- Robust log-based anomaly detection on unstable log data
- Anomaly Detection and Classification using Distributed Tracing and Deep Learning
- Demystifying the Black Box: A Classification Scheme for Interpretation and Visualization of Deep Intelligent Systems
- Detecting Anomalies From Big Data System Logs
- Visual explanation of neural network based rotation machinery anomaly detection system
- An effective spatial-temporal attention based neural network for traffic flow prediction
- Detecting Network Intrusions from Authentication Logs
- nLSALog: An Anomaly Detection Framework for Log Sequence in Security Management
- Detecting Anomalous Events on Distributed Systems Using Convolutional Neural Networks
- Software Logging for Machine Learning
- A Neural Attention Model for Real-Time Network Intrusion Detection
- Adversarial Discriminative Attention for Robust Anomaly Detection
Related papers
- DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning
- Application of deep learning to enhance the accuracy of intrusion detection in modern computer networks
- A Deep Learning Ensemble for Network Anomaly and Cyber-Attack Detection
- Explaining Network Intrusion Detection System Using Explainable AI Framework
- Intrusion Detection System Using Convolutional Neuronal Networks: A Cognitive Computing Approach for Anomaly Detection based on Deep Learning
- Performance Analysis of Machine Learning Techniques in Intrusion Detection