This project is a look into the potential of using Local Outlier Factors from Unsupervised Learning as a way of guiding exploration in a Reinforcement Learning Agent. The abstract of the resulting technical report reads:
We borrow the idea of Local Outlier Factors from Unsupervised Learning as a means of encouraging a Reinforcement Learning agent to explore uniformly. We show how this can be done and how it can be supplied to the agent as a simple reward signal. We also provide experimental results on a modified gridworld domain which gives strong evidence that this is a useful way of incentivizing uniform exploration. Finally, we discuss the key problems that have to be solved for this to be a practical method and how it can be further improved.
Included in this repository is the following:
- a technical report on the topic mentioned above
- a presentation that was given on the topic
- the LaTeX code used to compile the report
- the source code for a series of experiments elaborated on in the report which can be run by executing
run.shin an appropriate system
For any questions regarding this work email Dylan R. Ashley.