An On‐Line Approach for Classifying and Extracting Application Behavior on Linux
Luciano José Senger, Rodrigo Fernandes de Mello, Marcos José Santana, Regina Helena, Carlucci Santana, Laurence Tianruo Yang · 2005
This chapter contains sections titled: Introduction Related Work Information Acquisition Linux Process Classification Model Training algorithm Labeling Algorithm Classification Model Implementation Results Evaluation Of The Model Intrusion On The System Performance Conclusion Reference A new model for classifying and extracting the application behavior is presented on this article. The extraction and classification of process behavior are conducted through the analysis of data that represent the resource occupation. These data are captured and presented on an artificial self-organizing neural network, which is responsible for extracting the process behavior patterns. This architecture allows the periodic and on-line updating of the resource occupation data, which adapts itself to classify new behaviors of a same process. The proposed model aims to provide process behavior information for the load balancing algorithms that use this information on their selection and transference policies. The results obtained on experiments prove the contributions provided by the proposed model when compared to other works.