Week 6
Hello everyone, and welcome back to Week 6 of my Senior Research Project on Reinforcement Learning!
This week was all about brainstorming and defining the direction for the end product of our project. We focused on how to present our work in the most meaningful and impactful way, especially considering the complex nature of reinforcement learning and explainability. After discussing a few ideas, we realized that a key goal is to create a model that is not only effective but also interpretable. This led us to think about the explainability aspect in more depth and how we can clearly showcase it.
For inspiration, I spent some time exploring the TensorFlow Playground, a tool that lets you visualize how neural networks work with different datasets. The interactive platform shows how various parameters like activation functions, learning rates, and network shapes can affect the results. It’s a simple and engaging way to see how changes to a model’s design can impact its learning process. I’ve been trying to replicate something similar in our own project, but with a focus on explainable AI. The goal is to demonstrate not just how well our model performs but also why it makes certain decisions.
To make the model more explainable, we plan on using the CartPole environment to train the agent. By training our model in this simple yet informative environment, we hope to develop tools that allow us to visualize and explain how the agent is learning and making decisions in a way that’s understandable to users. Currently, we’re still in the early stages of this process, just working on getting everything set up and exploring how best to display the results.
I’m excited to continue working on this and sharing more as we make progress. Thank you for reading through my Week 6 update! If you have any questions, feel free to ask.
Until next time, I will see you later!
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.