Generating unit tests using reinforcement learning
Organization
Computational Neuroscience Lab
Abstract
Code coverage is an important metric for measuring the quality of unit tests. The idea of this project is to study if reinforcement learning could be used to improve code coverage. Recurrent neural network is used to parse code and generate function calls. Reinforcement learning is used to improve the network, taking number of lines covered as a reward. The student has to be familiar with neural networks for text processing and policy gradient method for reinforcement learning.
Graduation Theses defence year
2017-2018
Supervisor
Tambet Matiisen, Vesal Vojdani, Kristjan Sägi, Triin Kask
Spoken language (s)
Estonian, English
Requirements for candidates
Level
Masters
Keywords
Application of contact
Name
Tambet Matiisen
Phone
+3725286457
E-mail