personal website
Main menu:
Mikhail Kazdagli
Security Researcher
Email: mikhail.kazdagli@utexas.edu
Resume: Please e-mail me
References available upon request
Research Interests
- Computer Security and Software Reliability
- Static Code Analysis
- Machine Learning and Artificial Intelligence
Ongoing Projects
Currently, I am working on a Deep Learning-based framework for detecting an anomalous dynamic behavior in Android devices. Modern event-driven systems (e.g. Android OS) often suffer from transient performance issues, which are usually caused by an unpredictable resource contention (e.g. multiple callbacks are executed concurrently in the presence of external interrupts).
The current prototype uses an LSTM network that monitors an internal state of multiple Android components (e.g. system services, drivers) for abnormal behavior. The neural net comprises of several stacked LSTM layers and an output fully connected layer. It raises an alert if the distance between the LSTM’s prediction and the actual event is higher than a threshold. Read More...