Unsupervised real-time anomaly detection for streaming data.
Neurocomputing(2017)
摘要
•Real-world streaming analytics calls for novel algorithms that run online, and corresponding tools for evaluation.•Anomaly detection with Hierarchical Temporal Memory (HTM) is a state-of-the-art, online, unsupervised method.•The Numenta Anomaly Benchmark (NAB) is an open-source environment specifically designed to evaluate anomaly detection algorithms for real-world use.
更多查看译文
关键词
Anomaly detection,Hierarchical Temporal Memory,Streaming data,Unsupervised learning,Concept drift,Benchmark dataset
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络