Abstract:

Automated testing methods play an essential role in enhancing software security and development. Recent papers researching fuzzing concentrate on applications that leverage machine-learning approaches to overcome its primitive limitations. This article examines current research on the use of machine learning to automated fuzzing. It specifically examines fuzzer types, machine learning technological potential, and briefly discusses issues found in own research conducted. Future research to alleviate fuzzing bottlenecks is defined in this paper as the authors thoughts.