Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/duketheduck1/rldd

Reinforcement learning for Delta Debugging
https://github.com/duketheduck1/rldd

Last synced: 28 days ago
JSON representation

Reinforcement learning for Delta Debugging

Awesome Lists containing this project

README

        

# RLDD

## Project Description

This project is for COSC 4F90 at Brock University and focuses on the intersection of Reinforcement Learning (RL) and Delta Debugging—a pivotal area in software debugging. The primary aim is to find a novel approach to use delta debugging by incorporating RL technology.

### Project Objectives:

- **Integration of Reinforcement Learning:** Explore the use of RL techniques in the Delta Debugging methodology.

- **Automating Fault Isolation:** Develop an RL-driven model to automate the identification and isolation of minimal inputs responsible for triggering software bugs.

- **Efficiency Gains:** Investigate the potential efficiency gains of using newer technologies, contributing to enhanced debugging processes.

## TODO
- [x] Test with several libraries like Pytorch, TensorFlow, Scikit_learn
- [ ] Find a better reward system for Q learning
- [ ] Figure out how to work on multiple failure points using delta debugging
- [ ] Create instructions for installation since the whole project is just some "working" test file that I'm trying to work with Q-learning and delta debugging
- [ ] Explore strategies to extend the application of delta debugging to handle multiple failure points simultaneously, enhancing the algorithm's versatility.
- [ ] Identify additional features or improvements for the plugin that could enhance its functionality or user experience.
- [ ] Document and create detailed installation instructions to ensure a seamless onboarding experience for users interested in experimenting with Q-learning and delta debugging in this project.
- [ ] Conduct a thorough literature review to gather insights and best practices in combining Q-learning and delta debugging for fault isolation.

## License
This project is licensed under the MIT License