Title:
CLOUD NETWORK ABNORMALITY DETECTION MODEL TRAINING METHOD BASED ON REINFORCEMENT LEARNING, AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2024/104401
Kind Code:
A1
Abstract:
Embodiments of the present description provide a cloud network abnormality detection model training method based on reinforcement learning, and a storage medium. The method comprises: inputting, into a first detection network of an abnormality detection model, historical time sequence data related to a device to be detected for processing, to generate a first abnormality detection result corresponding to said device at at least one historical moment; determining a detection error corresponding to the first abnormality detection result, and inputting the detection error and at least one piece of historical alarm information of said device into an agent of a second detection network in the abnormality detection model for processing, to generate a second abnormality detection result of said device at a target moment; determining reward information of the second detection network according to the second abnormality detection result and a target abnormality detection result of said device at the target moment; and adjusting network parameters of the agent on the basis of the reward information, to generate a target abnormality detection model.
Inventors:
YANG JIAHAI (CN)
HE XIN (CN)
ZHANG SHIZE (CN)
DONG ENHUAN (CN)
WANG ZHILIANG (CN)
LU JIANYUAN (CN)
LYU BIAO (CN)
ZHU SHUNMIN (CN)
HE XIN (CN)
ZHANG SHIZE (CN)
DONG ENHUAN (CN)
WANG ZHILIANG (CN)
LU JIANYUAN (CN)
LYU BIAO (CN)
ZHU SHUNMIN (CN)
Application Number:
PCT/CN2023/131887
Publication Date:
May 23, 2024
Filing Date:
November 15, 2023
Export Citation:
Assignee:
HANGZHOU ALICLOUD FEITIAN INFORMATION TECH CO LTD (CN)
International Classes:
H04L41/0604
Foreign References:
CN115514614A | 2022-12-23 | |||
US20210097438A1 | 2021-04-01 | |||
CN113111996A | 2021-07-13 | |||
CN113743607A | 2021-12-03 | |||
CN114520743A | 2022-05-20 | |||
US20160062815A1 | 2016-03-03 | |||
US20210256392A1 | 2021-08-19 | |||
US20210406109A1 | 2021-12-30 |
Attorney, Agent or Firm:
GUANGZHOU ZHUZHI INTELLECTUAL PROPERTY LIMITED (CN)
Download PDF: