Random vs. Reflex Agents

In this project we implemented a random and a reflex agent in a trivial vacuum environment. The trivial vacuum environment only has two locations, and they can either be clean or dirty and the agent (a vacuum) in this environment only has four actions: go right, left, suck, or do nothing. The objective was to compare the agents' performace; since the reflex agent can perceive the environment it obtained a better result.

The agents were implemented using the UC Berkeley code repository, specifically the agents file. Here is our implementation for the assignment

Below is the report we prepared and a video of approximately five minutes summarizing our work.