Peng Lin

TL;DR: This paper design and implement a new anomaly detection platform - NetDetector for networked systems based on LSTM, a special kind of recurrent neural network, which achieves about 8% performance improvement as compared with systems that use traditional machine learning methods.

TL;DR: A method called DCSA (Density-based Clustering and Sequential Association Analysis) for alarm association rules mining using time density-based clustering and FP-Growth algorithm to mine the association rules in alarm data, which overcomes the drawbacks of sliding windows method.

TL;DR: In this paper, a network anomaly detection method and system, a terminal and a storage medium, is described, which comprises the following steps: carrying out vector conversion on network traffic by adopting an ngram model to obtain a vector matrix of the network traffic; performing spatial-temporal feature extraction on the vector matrix by adopting a long short-term memory network and a bidirectional gating cycle unit to obtain an implicit state of network traffic.

TL;DR: In this article, a telecommunication network alarm prediction method is proposed, which comprises the following steps: reading telecommunication networks alarm data in batches according to a time sequence and preprocessing the alarm data; carrying out DBSCAN clustering on the preprocessed alarm data according to the Unix timestamp; mining a frequent item set and an alarm association rule for the alarms after clustering; merging and storing alarm association rules mined from all batches of alarm data.