Anomaly detection in dynamic execution environments

摘要: In the past few years, languages which run on virtual machines, like Java and C#, have become popular. These are platforms as well languages, they characterized by being verifiable garbage collected, include Just-In-Time compilers, large standard libraries, runtime profilers. I call with these features dynamic execution environments (DEEs). The infrastructure of DEEs allows access to that were previously difficult observe. My research consists a series case studies in build systems classify behavior particular feature into normal abnormal then use classification for either security or optimization purposes. anomaly detectors. I detection method invocations, permissions, invocation sequences. them sandboxes, used detect intrusions system faults. also show an detector can be predict object lifetimes resulting improved collector.