SE4AI: Enabling Reliable Deep Learning via Static Code Analysis

There are two ways to improve the reliability of machine learning applications: (1) on the reliability of the machine learning model or algorithm and (2) on the reliability of the code implementing the application. This project will mainly focus on the latter case, for which our fellow researchers have not started exploiting it yet. This project hence aims at supporting developers to implement reliable machine learning applications, both at the development phase and release phase.