Master's thesis on generating test cases for web applications. The thesis proposes a test generation framework for test automation libraries.
Algorithms implemented for test generation:
- Vanilla Policy Gradient (PG)
- Proximal Policy Optimization (PPO)
- Online Decision Transformer (ODT)
Automation in web application testing improves efficiency and reduces costs. This thesis introduces a framework for generating test cases using machine learning algorithms, specifically PPO and ODT, to optimize test steps for user objectives. Results show the framework supports exploration-based test generation, with PPO optimizing steps and ODT cloning behavior from previous cases. The thesis also discusses scalability solutions and future goals for rapid, simultaneous test generation across applications.
Thesis: Lehtonen_2023.pdf
Archived project: GitHub Repository
