首页|Patent Issued for Application anomaly detection using trained infrastructure mac hine learning models (USPTO 12112209)
Patent Issued for Application anomaly detection using trained infrastructure mac hine learning models (USPTO 12112209)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Reporters obtained the following quote from the background information supplied by the inventors: “Application infrastructure consists of multi-layered componen ts required to deliver an application including its functions and services to a customer. However, the application infrastructure is susceptible to multiple app lication risks, such as infrastructure failure and security breaches. Thus, it is a challenging task to manage the application infrastructure. Conventionally, there are multiple systems available for managing the application infrastructure. However, the conventional systems generate a predictive recommendation model fo r managing the application infrastructure based on a single infrastructure compo nent’s behavior. Thus, the conventional systems fail to manage the application i nfrastructure efficiently. Further, the conventional systems fail to provide pat tern identification in the application infrastructure or customer infrastructure involving multiple related application infrastructure components forming the ap plication. Moreover, the conventional systems are unable to develop any knowledg e base models for prediction of anomalies, such as intrusion and faults, in appl ications by identifying the patterns in the application infrastructure. Furtherm ore, the conventional systems also fail to develop the knowledge base models for prediction of the anomalies in the applications by correlation-learning from on e application’s behavior to another application with a similar infrastructure.
NETL
NSTL
