Title:
ANOMALY DETECTION METHOD FOR LARGE-SCALE MULTIVARIATE TIME SERIES DATA IN CLOUD ENVIRONMENT
Document Type and Number:
WIPO Patent Application WO/2022/160902
Kind Code:
A1
Abstract:
An anomaly detection method for large-scale multivariate time series data in a cloud environment. The method comprises: establishing an anomaly detection model for multivariate time series data by means of offline training, and performing anomaly detection on online monitored data by means of the offline-trained anomaly detection model. According to the method, the feedforward neural network of a native variational autoencoder is improved in the stage of offline model training to construct the dependency of multivariate time series; a loss function calculation method is improved, so that during model training, the data in a normal mode can be paid attention to and the data in an abnormal mode can be ignored, so that when an anomaly occurs during online anomaly detection, the probability of model reconstruction is low, and it is easier to detect the anomaly.
Inventors:
CHEN NINGJIANG (CN)
DUAN XIAOYAN (CN)
LIU KANGKANG (CN)
DUAN XIAOYAN (CN)
LIU KANGKANG (CN)
Application Number:
PCT/CN2021/133024
Publication Date:
August 04, 2022
Filing Date:
November 25, 2021
Export Citation:
Assignee:
UNIV GUANGXI (CN)
International Classes:
G06N3/04
Foreign References:
| CN112784965A | 2021-05-11 | |||
| CN109492193A | 2019-03-19 | |||
| CN111858231A | 2020-10-30 | |||
| CN112131212A | 2020-12-25 | |||
| CN112163020A | 2021-01-01 | |||
| US20200097810A1 | 2020-03-26 |
Attorney, Agent or Firm:
LIUJIA CHINA IP LAW OFFICE (CN)
Download PDF: